WHEELCHAIR WITH MULTI-AXIS PASSIVE-ACTIVE SUSPENSION

ACTUATION

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to and the benefit of the filing date of U.S. Provisional Patent Application No. 63/188,591, filed May 14, 2021, the entirety of which is hereby incorporated by reference herein.

FIELD

[0002] This application relates to personal mobility devices and, in particular, to suspension systems of wheelchairs.

STATEMENT OF GOVERNMENT SUPPORT

[0003] This invention was made with government support through grants B2988C, A3076M, and B9269-L of the U.S. Department of Veterans Affairs. The government has certain rights in the invention.

BACKGROUND

[0004] Wheeled power mobility is critical to allowing individuals with mobility impairments to participate in life activities and advance their lives. The ability to negotiate indoor and outdoor environments, and the obstacles that those environments present, is critical to achieving effective functional mobility. To effectively negotiate the obstacles that are frequently encountered during use of personal powered mobility devices, it is important to be able to smoothly navigate irregular and uneven surfaces and overcome curbs, slopes, cross slopes, dips, and bumps in a safe and comfortable manner.

SUMMARY

[0005] Disclosed herein, in one aspect, is a wheelchair comprising a frame coupled to a seat and a pluralit of wheel assemblies. Each wheel assembly can comprise a swing arm that is pivotably coupled to the frame, a wheel rotatably coupled to the swing arm about a respective rotational axis, and an actuator system that is configured to adjust the position of the wheel relative to the frame. The actuator system (e g., a passive-active actuator system) can comprise an actively controlled actuator and a passive shock absorber arranged in series with the actively controlled actuator. At least one force sensor can be configured to provide an

1 output associated with a ground force applied by the ground to the wheel. At least one position sensor can be configured to provide an output associated with a position of the wheel. The wheelchair can further comprise a memory storing therein at least one center of mass threshold and/or movement trajectory.

[0006] At least one processor can be in communication with the memory and in communication with the force sensor(s) and the position sensor(s) of each wheel assembly. The at least one processor can be configured to: determine, based on the outputs of the force sensor(s) and position sensor(s) of each wheel assembly of the plurality of wheel assemblies, a center of mass of the wheelchair; compare the center of mass of the wheelchair to the at least one center of mass threshold; and adjust, using the actively controlled actuator of at least one wheel assembly of the plurality of wheel assemblies, the center of mass of the wheelchair if the center of mass exceeds the at least one threshold.

[0007] Additional advantages of the disclosed system and method will be set forth in part in the description which follows, and in part will be understood from the description, or may be learned by practice of the disclosed system and method. The advantages of the disclosed system and method will be realized and attained by means of the elements and combinations particularly pointed out in the appended claims. It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention as claimed.

BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate several embodiments of the disclosed apparatus, system, and method and together with the description, serve to explain the principles of the disclosed apparatus, system, and method.

[0009] FIG. 1 is a perspective view of a wheelchair in accordance with embodiments disclosed herein.

[0010] FIG. 2A shows a side view of a wheel assembly of the wheelchair of FIG. 1. FIG. 2B shows a side perspective view of the wheel assembly of the wheelchair of FIG. 1.

2 [0011] FIG. 3 A shows a rear perspective view of the wheel assembly of FIGS. 2A-2B. FIG. 3B shows a rear view of the wheel assembly of FIGS. 2A-2B. The large circles provided on the drawings represent pivot points.

[0012] FIG. 4 shows a top view of a wheel assembly for a drive wheel of the wheelchair of FIG. 1.

[0013] FIG. 5 shows a bottom view of the wheel assembly for the drive wheel of FIG. 4.

[0014] FIG. 6 shows a side perspective view of the wheel assembly for the drive wheel of FIG. 4.

[0015] FIG. 7 shows a block diagram of an exemplary control loop for controlling the movement of each of the wheel assemblies as disclosed herein.

[0016] FIG. 8A shows a schematic diagram of a center of mass in relation to center of mass thresholds. FIG. 8B shows a schematic diagram of a cross-section of an ellipsoid-shaped threshold at a first height. FIG. 8C shows a schematic diagram of a cross-section of the ellipsoid-shaped threshold at a second height.

[0017] FIG. 9 shows a block diagram of sensors and controls of the wheelchair as disclosed herein.

[0018] FIG. 10 illustrates a computing system comprising a computing device in accordance with embodiments disclosed herein.

[0019] FIG. 11 A illustrates a first step in climbing an obstacle, such as a curb. FIG. 1 IB illustrates a second step in climbing the obstacle. FIG. 11C illustrates a third step in climbing the obstacle. FIG. 1 ID illustrates a fourth step in climbing the obstacle. FIG. 11E illustrates a fifth step in climbing the obstacle. FIG. 1 IF illustrates a sixth step in climbing the obstacle.

[0020] FIG. 12A illustrates a first step in descending an obstacle, such as a curb. FIG. 12B illustrates a second step in descending the obstacle. FIG. 12C illustrates a third step in descending the obstacle. FIG. 12D illustrates a fourth step in descending the obstacle. FIG. 12E illustrates a fifth step in descending the obstacle. FIG. 12F illustrates a sixth step in descending the obstacle.

3 [0021] FIG. 13A depicts an exemplary machine vision sensor system. FIG. 13B shows a sensed area and calculated data associated with the sensed area, the calculated data usable for traversing the sensed area and which can be provided on a display.

[0022] FIG. 14A illustrates a schematic diagram for an exemplary wheelchair as disclosed herein. FIG. 14B shows a schematic diagram for a wheelchair without active actuation.

[0023] FIG. 15 shows four obstacles traversed by an exemplary wheelchair as disclosed herein.

[0024] FIGS. 16A-16B show respective charts of root mean square (RMS) and vibration dose values (VDV) for 1) a conventional electric wheelchair, 2) a wheelchair as disclosed herein without active suspension, and 3) a wheelchair as disclosed herein with active suspension across different obstacles.

[0025] FIGS. 17A-17B show respective charts of root mean square (RMS) and vibration dose values (VDV) for 1) a conventional electric wheelchair, 2) a wheelchair as disclosed herein without active suspension, and 3) a wheelchair as disclosed herein with active suspension across different obstacles.

[0026] FIG. 18 is a schematic showing input devices and user interface screens for operating an exemplary wheelchair as disclosed herein.

[0027] FIG. 19 shows an exemplary wheelchair in a raised configuration.

[0028] FIG. 20 shows a flow chart of an algorithm for ascending and descending an obstacle with the exemplary wheelchair as disclosed herein.

[0029] FIG. 21 is a simplified schematic diagram illustrating the linkage of a drive wheel and carriage. As shown, larger circles indicate pivot points that are not movable relative to the carriage, and smaller circles indicate pivot points that are movable relative to the carriage.

DETAILED DESCRIPTION

[0030] The disclosed system and method may be understood more readily by reference to the following detailed description of particular embodiments and the examples included therein and to the Figures and their previous and following description.

4 [0031] It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of the present invention which will be limited only by the appended claims.

[0032] It must be noted that as used herein and in the appended claims, the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise. Thus, for example, reference to a “wheel assembly” includes one or more of such wheel assemblies, and so forth.

[0033] “Optional” or “optionally” means that the subsequently described event, circumstance, or material may or may not occur or be present, and that the description includes instances where the event, circumstance, or material occurs or is present and instances where it does not occur or is not present.

[0034] Ranges may be expressed herein as from “about” one particular value, and/or to “about” another particular value. When such a range is expressed, also specifically contemplated and considered disclosed is the range from the one particular value and/or to the other particular value unless the context specifically indicates otherwise. Similarly, when values are expressed as approximations, by use of the antecedent “about,” it will be understood that the particular value forms another, specifically contemplated embodiment that should be considered disclosed unless the context specifically indicates otherwise. It will be further understood that the endpoints of each of the ranges are significant both in relation to the other endpoint, and independently of the other endpoint unless the context specifically indicates otherwise. Finally, it should be understood that all of the individual values and sub ranges of values contained within an explicitly disclosed range are also specifically contemplated and should be considered disclosed unless the context specifically indicates otherwise. The foregoing applies regardless of whether in particular cases some or all of these embodiments are explicitly disclosed.

[0035] Optionally, in some aspects, when values or characteristics are approximated by use of the antecedents “about,” “substantially,” or “generally,” it is contemplated that values within up to 15%, up to 10%, up to 5%, or up to 1% (above or below) of the particularly stated value or characteristic can be included within the scope of those aspects.

[0036] Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of skill in the art to which the disclosed apparatus,

5 system, and method belong. Although any apparatus, systems, and methods and materials similar or equivalent to those descnbed herein can be used in the practice or testing of the present apparatus, system, and method, the particularly useful methods, devices, systems, and materials are as described.

[0037] Throughout the description and claims of this specification, the word “comprise” and variations of the word, such as “comprising” and “comprises,” means “including but not limited to,” and is not intended to exclude, for example, other additives, components, integers or steps. In particular, in methods stated as comprising one or more steps or operations it is specifically contemplated that each step comprises what is listed (unless that step includes a limiting term such as “consisting of’), meaning that each step is not intended to exclude, for example, other additives, components, integers or steps that are not listed in the step.

[0038] As used herein, the term “caster” should be understood to include both conventional casters (e.g., wheels that swivel about an axis that is perpendicular to their axis of rotation) as well as omni-wheels, which include rollers that permit movement perpendicular to the axis of rotation of the wheel.

[0039] Disclosed herein, in vanous aspects and with reference to FIG. 1 is a wheelchair 10 having a seat 12 and a frame 14 coupled to the seat. The wheelchair 10 can have a longitudinal axis 6 that extends between a front and a rear of the wheelchair 10 and a transverse axis 8 that is perpendicular to the longitudinal axis and extends between left and right sides of the wheelchair.

[0040] A plurality of wheel assemblies 20 can couple to the frame 14 so that the plurality of wheel assemblies distribute the weight of the wheelchair on the ground. For example, the plurality of wheel assemblies 20 can comprise a pair of drive wheel assemblies 22, with the two drive wheel assemblies of the pair being positioned on opposing sides of the frame 14. The plurality of wheel assemblies 20 can further comprise at least one freely rotating wheel assembly 24. The freely rotating wheel assembly (or assemblies) 24 can optionally be positioned rearwardly of the drive wheel assemblies 22 along the longitudinal axis 6. In some aspects, the at least one freely rotating wheel assembly 24 can consist of a single freely rotating wheel assembly that is positioned rearwardly of the drive wheel assemblies 22 along the longitudinal axis 6. In various aspects, the wheelchair can further comprise freely rotating front wheels 100.

6 [0041] Referring also to FIGS. 2-3, the freely rotating wheel assembly 24 can comprise a swing arm 30 that is pivotably coupled to the frame 14 about a pivot axis 31. At least one wheel 32 can be coupled to the swing arm 30 and rotatable about a rotational axis 34. An actuator system 36 can be configured to adjust the position of the swing arm 30 relative to the frame, thereby adjusting a vertical position of the wheel relative to the frame 14. The actuator system 36 can comprise an actively controlled actuator 38 and a passive shock absorber 40 that is coupled to the actively controlled actuator 38 in series. That is, the actively controlled actuator 38 and the passive shock absorber 40 can each independently affect the vertical position of the wheel to which the actuators are coupled. Accordingly, each of the wheels can move in response to the driving surface through the shock absorber or in response to the movement of the active actuator.

[0042] For example, in some aspects, the passive shock absorbers 40 can be linear shock absorbers having a first end 42 and a second end 44, wherein a biasing element (not shown, but as is conventionally provided within linear shock absorbers) can bias the first end away from the second end. The biasing element can be, for example, a coil spring or a gas spring or both. The first end 42 can couple to the swing arm 30, and the second end 44 can be coupled to the actively controlled actuator 38 via a linkage 46. In this way, actuation of the actively controlled actuator can determine a position of the first end of the passive shock absorber, and the biasing element of the passive shock absorber can allow movement of the second end of the passive shock absorber relative to the first end to provide passive suspension of the wheel(s) 32. The wheelchair disclosed herein can be driven via an on board power supply (e.g., a battery 116 - FIG. 9). It is contemplated that use of a passive shock absorber can significantly reduce an energy load on the power source as compared to using an active suspension to absorb small perturbations.

[0043] In some aspects, the linkage 46 can comprise a first linkage element 48 that is pivotably coupled to the frame 14 about a pivot axis 50, wherein a first end 52 of the first linkage element is coupled to the actively controlled actuator 38 (e.g., via a piston rod 54 of the actively controlled actuator). A second end 56 of the first linkage element 48 can couple to a first end 58 of a second linkage element 60. A third linkage element 62 can couple to the frame 14 at a first end 64. A second end 66 of the second linkage element 60 can couple to a second end 68 of the third linkage element 62. The second end 68 of the third linkage element 62 can couple to the second end of the passive shock absorber 40. Thus, the frame

7 14 and the first, second, and third linkage elements 48, 56, 62 can define a four-bar linkage that can adjust the second end of the passive shock absorber 40.

[0044] Optionally, the freely rotating wheel assembly 24 can comprise a pair of wheels 32 that are spaced along the transverse axis 8 on an axle 33. Optionally, the wheels 32 can couple to the swing arm 30 via a torsion joint 69 that can enable the wheels to pivot about a torsion axis 71 (that can optionally be parallel to, or coaxial with, the longitudinal axis 6).

The torsion joint 69 can be actively or passively controlled. For example, a passive biasing element 73 can bias the torsion joint toward a position in which a rotational axis 34 of the wheels is parallel to the transverse axis 8. In some optional aspects, the passive biasing element 73 can comprise an elastomeric material (e.g., polyurethane). Optionally, the passive biasing element 73 can comprise first and second components on each side of the torsion axis 71. For example, the first and second components can be received between blind holes in each of the swing arm 30 and the axle 33 on either side of the torsion axis 71. Thus, pivotal movement of the axle 33 relative to the swing arm 30 about the torsion axis 71 can apply a shear force to the elastomeric material. Optionally, the elastomeric material can comprise a rigid element (e g., a steel rod) that extends through the elastomeric material (optionally , through the center of the elastomeric material). The rigid element can retain the elastomeric material in compression to inhibit tearing or shearing of the elastomeric material. In further aspects, a rotational actuator can actively adjust the rotational axis of the wheels 34 about the torsion axis 71. The actuator can be configured to actively maintain both of the wheels 32 in contact with the ground surface.

[0045] Referring to FIGS. 4-6, each of the drive wheel assemblies 22 can comprise a swing arm 70 that is pivotably coupled to the frame 14 about a pivot axis 72. A drive wheel 74 can be rotatably coupled to the swing arm at a first end 80 about a rotational axis 76. A drive motor 78 can be coupled to the drive wheel 74 and configured to cause the drive wheel to rotate. An actuator sy stem 82 can be configured to adjust the vertical position of the drive wheel 72 relative to the frame 14. In some optional aspects, the actuator system 82 can pivotably couple to the swing arm 70 at a pivotable coupling 83 at a second end of the swing arm with the pivot axis 72 between the coupling of the drive wheel 74 to the swing arm and the coupling between the actuation sy stem 82 and the swing arm.

[0046] Referring also to FIG. 21, the actuator system 82 of each drive wheel assembly 22 can comprise an actively controlled actuator 84 coupled in series to a passive shock absorber 86

8 (e.g., a linear shock absorber). For example, in some aspects, the actively controlled actuator 84 can be a linear actuator that has a first end that is pivotably coupled to the swing arm 70 at the pivotable coupling 83. The actively controlled actuator 84 can comprise a piston rod 88 that is parallel or generally parallel to (e.g., within 15 degrees of, 10 degrees of, or 5 degrees of, or 1 degree of) the longitudinal axis 6 of the wheelchair. The passive shock absorber 86 can have a first end 90 that is movable (e.g., axially movable) relative to a second end 92.

The first end 90 of the passive shock absorber 86 can couple to the carriage 95 at a pivot point 87. The piston rod 88 can be coupled to the second end 92 of the passive shock absorber 86. For example, a linkage 94 (e.g., a four-bar linkage) can couple the piston rod 88 of the actively controlled actuator 84 to the passive shock absorber 86. The linkage 94 can further set or establish a mechanical advantage between the movement of the passive shock absorber and the actively controlled actuator 84. In some aspects, the linkage 94 can be configured so that movement of the piston rod 88 in a first direction can cause movement of the second end 92 of the passive shock absorber 86 in a second direction that is opposite or generally opposite the first direction (e.g., as shown and described for the freely rotating wheel assembly 24). For example, forw ard movement of the piston rod 88 can cause rearward movement of the second end of the passive shock absorber. This configuration can provide for a compact configuration. In exemplar aspects, it is contemplated that each of the actively controlled actuator and the passive shock absorber can be elongate in the respective dimension in which it extends and retracts. Accordingly, in providing a linkage as described herein, the actively controlled actuator and the passive shock absorber can be positioned alongside each other. The same advantage can be appreciated with the configuration of the freely rotating wheel assembly 24.

[0047] In some aspects, the active actuators 38, 84 can be electrohydraulic actuators. In these aspects, one or more electric motors 96 can be configured to provide hydraulic power to the hydraulic actuators.

[0048] Each of the wheel assemblies 20 can comprise a force sensor 102 that is configured to determine a ground force applied by the ground to the respective wheel. In various aspects, it is contemplated that the force sensor 102 can be positioned anywhere within the coupling between the wheel and the frame so that the weight against the wheel causes a proportional force against the force sensor. For example, the force sensor 102 of the drive wheel assembly 22 can be positioned between the linkage 94 and the swing arm 70 (e.g., between the swing

9 arm 70 and the actively controlled actuator 84, as illustrated). Based on geometric ratios and mechanical advantages of couplings between linkages, the force measured by force sensor 102 (e.g., a load cell) can be converted to the ground force experienced by the respective wheel.

[0049] Each of the wheel assemblies 20 can further comprise a position sensor 104 that is configured to provide an output associated with a position of the respective wheel. For example, the position sensor 104 can be a rotary sensor that is configured to detect a rotational position of the swing arm 30,70 (of a respective wheel assembly 20) that can be used to determine a position of the wheel relative to the frame (e.g., a vertical position of the wheel relative to the frame). In further aspects, the position sensor 104 can be a linear sensor. For example, the linear sensor can be coupled to the swing arm and the frame at respective fixed points. Using geometry of the triangle defined by the hinge point of the swing arm and the fixed points, the angle of the swing arm can be determined. In various aspects, the position sensors 104 can be linear or rotary encoders, potentiometers, and/or LVDs.

[0050] The wheelchair 10 can further comprise one or more orientation sensors 106. The wheelchair 10 can further comprise one or more acceleration sensors 108 (e.g., accelerometers). The acceleration sensors 108 can be configured to detect vibration (e.g., amplitude and/or frequency). In further aspects, the acceleration sensors 108 can be configured to detect shock. Still further, the accelerometers can be configured to determine speed and orientation (e.g., using numerical integration of acceleration measurements).

[0051] The wheelchair 10 can further comprise one or more speed sensors 110. In some optional aspects, the speed sensors(s) 110 can be respective encoders that are in communication with each drive wheel 74. The speed sensor(s) 110 can determine speed, for example, angular velocity of the drive wheels, to thereby determine linear speed of the wheelchair. In some optional aspects, the wheelchair can, based on feedback from the speed sensors 110, regulate a top speed of the wheelchair. It is contemplated that the speed sensors can be used in conjunction with a computing device and various other sensors, further described herein, to determine a maximum speed for a particular circumstance or mode in order to limit overspeeding of the wheelchair. For example, the computing device is configured to regulate a maximum speed of the wheelchair based on a proximity of the center of mass to at least one threshold as further disclosed herein. In further aspects, the maximum speed can be limited based on a quality of the terrain (e.g., bumpiness as measured by the

10 acceleration sensors 108). In still further aspects, the wheelchair can have limited speed based on a mode such as, for example, during self-leveling, during curb ascending/descending, etc.

[0052] In still further aspects, the wheelchair can comprise at least one machine vision sensor 112 (see also FIG. 13 A) such as, for example, one or more cameras, one or more laser range finders, and/or RADAR/LIDAR. For example, the cameras, LIDAR, and/or RADAR can be used in conjunction with a computing device, further disclosed herein, to detect and characterize obstacles for path determination. For example, in optional aspects, the machine vision sensor(s) 112 can be used to detect a curb that the wheelchair is configured to traverse. In further aspects, the machine vision sensor(s) 112 can be used detect a ramp (or other alternative path) to avoid a curb, stairs, or other obstacle. Still further, the machine vision sensor(s) 112 can assist in accelerating negotiation of various terrain.

[0053] The wheelchair 10 can comprise a computing device 1001 (FIG. 10) that can be in communication with the various sensors (e.g., force sensors 102, position sensors 104, orientation sensor(s) 106, acceleration sensors 108, speed sensors 110, and/or machine vision sensors 112) of the wheelchair. The computing device 1001 can be configured to control, based on feedback from the various sensors, various aspects of the wheelchair 10, including at least some (optionally, all) of the following: regulation of the seat orientation and attitude, obstacles detection and classification, path planning with flexible execution (i.e., path for curb climbing is statically stable and reversible at any point along the execution path), combined passive-active suspension for shock and vibration suppression.

[0054] Referring to FIGS. 7 and 8A-8C, the computing device 1001 (FIG. 10) can use one or more feedback loops that control the speed and direction of each vertical wheel movement and activation of its actuators to maintain a Center of Mass 118 (COM) within safety boundaries. For example, the computing device can have a memory 1012 that stores one or more center of mass thresholds 120. As shown in FIG. 8A, in some aspects, the center of mass thresholds can be a forward threshold 120a, a rearward threshold 120b, a left-side threshold 120c and a right-side threshold 120d. In further aspects, the center of mass threshold 120 can be defined by at least a portion of an ellipsoid. For example, FIG. 8B shows a cross section of the center of mass threshold ellipsoid, taken at a first height from the ground, and FIG. 8C shows a cross section of the center of mass threshold ellipsoid, taken from a second height that is greater than the first height. The ellipsoid can have a center that corresponds to the center of mass when the user is seated on the wheelchair and

11 the wheelchair is on flat, level ground. Accordingly, in some aspects, the higher the center of mass of the wheelchair is, the less tilt can be allowable. Optionally, the center of mass threshold can be calibrated on a reference surface with the user in the wheelchair. In some aspects, the center of mass threshold ellipsoid (or other threshold boundary) can be scaled by a safety factor from a maximum- tolerance center of mass threshold ellipsoid (or other threshold boundary). For example, the center of mass tolerance can be scaled as a fraction of maximum tolerance of the safety factor (e.g., (maximum tolerance)/ (safety factor)). Thus, for example, in some optional aspects, a safety factor of 2 can set the center of mass tolerances to half of the distance between the origin and the maximum tolerance center of mass threshold ellipses. Accordingly, the wheelchair can be configured to maintain the center of mass within 50% of the end point of the stability limit of each axis of the ellipsoid.

[0055] In some aspects, the threshold 120 can be modified based on the movement of the passive shock absorbers 40. For example, the wheelchair can increase the safety factor in response to detecting excessively bumpy terrain (e.g., as sensed by the accelerometers 108 or a sensor associated with the passive shock absorbers). In further aspects, upon sensing excessively bumpy terrain (or in response to the center of mass getting too close to the threshold 120 in the absence of excessively bumpy terrain) the actively controlled actuators 38 can move to adjust the center of mass away from a closest portion of threshold 120.

[0056] Referring to FIG. 7, the computing device 1001 (FIG. 10) can begin by calculating a Center of Mass (COM) Reference Model, via each wheel arm (w, /), using Vertical Force (Fzw, i ) and Joint Angle (Ww, i ) that includes both the user and wheelchair mass. The COM reference model can be used to determine the desired wheel arm vertical forces. The computing device can use said desired wheel arm vertical forces for low-level control. The low-level control can determine the speed and direction of each wheel arm based on the desired vertical force and estimated vertical position (I.w) and acceleration (tw) of each wheel arm. The latter two variables can be estimated through the powered personal wheeled mobility device inverse kinematics. These variables can be provided to an impedance controller that measures the difference (error) between the desired and estimated vertical forces and position to then obtain each wheel arm desired position and speed using proportional -integral-derivative (PID) control. Additionally, a gain scheduling system can affects the direction and activation of each wheel based on threshold values set by the seat angles (pitch (Oscai) and roll (bk»)) and vertical acceleration (vibration (<¾)) of the seat. For example, if a _z is higher than a predetermined vibration threshold but the seat angles (or center of mass) are within

12 the safety threshold, then the actively controlled (swing arm) actuators 38,84 can be inhibited to offer a passive suspension via the shock absorbers; however, if the seat angles (or center of mass) are over their safety threshold, regardless of the vibration, then the actuators can be engaged to maintain the seat orientation and attitude. In this way, the wheelchair can avoid excessive actuator movement that can dram power or exhaust actuator duty cycle. To complete the feedback loop, the COM location is calculated with powered personal wheeled mobility device dynamics to compare the reference model to the current state of the wheelchair.

[0057] In this way, advantageously, the disclosed wheelchair can automatically, based on sensed parameters, alternate between position control, in which the active actuators 38 maintain the center of mass within the threshold(s) 120, and force control, in which the actuators actively accommodate ground forces of the wheelchair to keep the wheelchair stable and improve the comfort of the user.

[0058] It is contemplated that determining the center of mass of the wheelchair as described herein can, by using ground force measurements, account for the weight of the user as well as the seated position of the user. This can be superior to actively controlled suspensions systems that neglect the weight of the user.

[0059] In various aspects, it is contemplated that the computing device can inhibit actuation of the passive shock absorbers when the COM is outside of the thresholds. For example, to inhibit tipping of the wheelchair, the computing device can, based on the center of mass being outside of at least one threshold, lock the passive shock absorbers to inhibit further movement of the COM away from the threshold. It is contemplated that the passive shock absorbers 40 can comprise sensors that are configured to detect rate and/or amount of compression.

[0060] In still further aspects, it is contemplated that the passive shock absorbers can have variable dampening impedance. For example, the passive shock absorbers can be gas springs in which movement of the gas spring causes fluid to pass through an orifice. It is contemplated that the size of the orifice can be adjusted in order to change the dampening impedance of the actuators. In this way, the impedance dampening can be tuned to different frequencies and loads. Thus, the computing device can adjust the impedance dampening of the passive shock absorber based on the vibration frequency detected by the acceleration sensors 108. Accordingly, in some aspects, to effect changes in impedance dampening, the shock absorbers can comprise an electro-hydraulic valve that is configured to change

13 impedance damping. In further aspects, the wheelchair can comprise pneumatic actuators that use an air source (e.g., a pump or a compressed gas cylinder) to adjust the impedance of the passive shock absorber.

[0061] It is known that determining the weight of a wheelchair user can be difficult. It is contemplated that the ground force measurements can be used to determine a weight of the user in the wheelchair. For example, the computing device can sum the ground forces of all of the wheel assemblies 20, subtract the weight of the wheelchair 10, and output the difference (e.g., to a display device 1011 (FIG. 10)).

[0062] In some aspects, the drive wheel assemblies 22 can be movable forwardly and rearwardly along the longitudinal axis relative to the frame 14 of the wheelchair 10. For example, and with reference to FIGS. 19 and 21, the drive wheel assemblies 22 can couple to the frame by respective carriages 95 that are movable forwardly and rearwardly along the frame 14. In exemplary aspects, the carriages 95 can move along one or more tracks 97 (FIGS. 1 IB and 11C). In exemplary aspects, at least one track on each side of the frame can comprise a rack 98 defining teeth. An actuator (e.g., an electric motor) can rotate a pinion against the rack 98 to move the carriage forwardly and rearwardly. In this way, the drive wheels can be positioned to optimally distribute load for balance. Still further, the forward and rearward movement of the drive wheels can enable traversing of obstacles as further described herein.

[0063] In various aspects, the computing device can cause the wheelchair to climb over an obstacle. Further details for obstacle climbing are described in U.S. Patent No. 10,912,688 (the ‘688 Patent), granted May 9, 2021, the entirety of which is hereby incorporated by reference herein. It should be understood that the wheels 32 can be raised and lowered via a single actively controlled actuator 38. Similarly, it is contemplated that the front casters 100 can be raised via a single actuator 130. Accordingly, it is contemplated that the methods for climbing a curb described in the ‘688 Patent can be consistent with operation of the wheelchair 10 described herein. Moreover, in optional aspects, a single actuator can raise and lower the wheels 32, and a single actuator can raise and lower the front casters 100. Further details of obstacle navigation are disclosed herein in Example 1 and with reference to FIGS. 11-12.

14 [0064] In various aspects, the computing device can, in conjunction with the machine vision sensors 112, perform obstacle detection and classification. The computing device can further perform path planning with flexible execution. For example, a path for curb climbing can be used as it is statically stable and reversible at any point along the execution path). According to at least one aspect, the machine vision sensors 112 can detect an obstacle (e.g., a curb).

The computing device 1000 can determine one or more proposed paths. For example, a first proposed path can be circumventing the obstacle. A second proposed path can be, for example, a preferred path to climb the obstacle. For example, the path can include aligning the wheelchair with the longitudinal axis 6 orthogonal (or within 5 degrees or 10 degrees of horizontal) to the curb. The computing device 1000 can further verify that the curb height is within the capabilities of the wheelchair. The computing device 1000 can then display the proposed paths on the display device 1011 (optionally, providing a recommended option).

The computing device can receive, from the user, an input confirming a proposed path or a selection from the alternatives of the proposed paths. Upon receiving said user input, the computing device can control movement of the actuator. In some aspects, during traversing of the selected path, the wheelchair can receive, from the user, an input device (e.g., via a joystick), allowing the user to move the wheelchair forward along the path, reverse movement along the path, or pause movement along the path. It is contemplated that feedback from the speed sensors 110 can be used in maintaining desired speeds of the wheelchair as it ascends or descends curbs. That is, the speed sensors can be used in speed control. For example, in curb climbing/descending, speed control can be used in approaching the curb to inhibit collision with, or falling over, the curb. The speed control can further be used to slowly rotate the drive wheels as the drive wheels contact the curb to draw the wheelchair forwardly upon contact. The speed control can further be used to move the wheelchair forwardly once the wheelchair has ascended/descended the curb.

[0065] The wheelchair as disclosed herein can provide several advantages over conventional powered wheelchairs. For example, the wheelchair as disclosed herein can sense ground reaction forces exerted on the wheels (e.g., the caster wheels and the drive wheels). Optionally, each of the wheels (drive wheels, front wheels, and rear wheels) can be in engagement with the ground, and the total ground reaction forces, can be summed to determine the total ground reaction force between the wheelchair and the ground. Using the ground reaction forces and position sensors, the wheelchair can determine a center of mass of both the wheelchair and the user, collectively. The wheelchair can determine the orientation

15 of the wheelchair and adjust the wheel positions to maintain the seated stability of the user (e.g., maintaining the pitch and roll of the seat of the wheelchair within respective thresholds). The wheelchair can comprise passive shock absorbers that suppress small perturbations of the wheels for a more comfortable ride than conventional wheelchairs. The wheelchair can comprise actively controlled actuators that are configured to adjust the position of the wheels of the wheelchair, and said actively controlled actuators can enable negotiation of large obstacles, such as, for example, curbs, slopes, cross-slopes, speed bumps, and potholes. The passive shock absorbers can be in series with the actively controlled actuators. The wheelchair can comprise a compliment of sensors that can include, for example, one or more of the following: ground-reaction force sensor(s), wheel/caster position sensor(s), seat/frame orientation sensor(s), seat/frame position sensor(s), seat/frame acceleration (i.e., vibration) sensor(s), machine vision (e.g., cameras, laser range finder, and RADAR) sensor(s), and/or wheel speed/acceleration sensor(s). In combination with the passive-active suspension for shock and vibration suppression, the wheelchair can further include a distributed control system (e.g., a computing device) that regulates the seat orientation and attitude, performs obstacle detection and classification, path planning with flexible execution (e.g., path planning for curb climbing is statically stable and reversible at any point along the execution path).

Computing Device

[0066] FIG. 10 shows a system 1000 including an exemplary configuration of a computing device 1001 for use with the wheelchair 10. In some aspects, the computing device 1001 can be integral to the wheelchair 10. In further aspects, it is contemplated that a separate computing device, such as, for example, a tablet, smartphone, laptop, or desktop computer can communicate with the wheelchair 10 and can enable the user to interface with the wheelchair 10.

[0067] The computing device 1001 may comprise one or more processors 1003, a system memory 1012, and a bus 1013 that couples various components of the computing device 1001 including the one or more processors 1003 to the system memory 1012. In the case of multiple processors 1003, the computing device 1001 may utilize parallel computing.

16 [0068] The bus 1013 may comprise one or more of several possible types of bus structures, such as a memory bus, memory controller, a peripheral bus, an accelerated graphics port, and a processor or local bus using any of a variety of bus architectures.

[0069] The computing device 1001 may operate on and/or comprise a variety of computer readable media (e g., non-transitory). Computer readable media may be any available media that is accessible by the computing device 1001 and comprises, non-transitory, volatile and/or non-volatile media, removable and non-removable media. The system memory 1012 has computer readable media in the form of volatile memory, such as random access memory (RAM), and/or non-volatile memory, such as read only memory (ROM). The system memory 1012 may store data such as sensor data 1007 (i.e., data from signals received by the electrodes) and/or program modules such as operating system 1005 and center of mass determination software 1006 that are accessible to and/or are operated on by the one or more processors 1003.

[0070] The computing device 1001 may also comprise other removable/non-removable, volatile/non-volatile computer storage media. The mass storage device 1004 may provide non-volatile storage of computer code, computer readable instructions, data structures, program modules, and other data for the computing device 1001. The mass storage device 1004 may be a hard disk, a removable magnetic disk, a removable optical disk, magnetic cassettes or other magnetic storage devices, flash memory cards, CD-ROM, digital versatile disks (DVD) or other optical storage, random access memories (RAM), read only memories (ROM), electrically erasable programmable read-only memory (EEPROM), and the like.

[0071] Any number of program modules may be stored on the mass storage device 1004. An operating system 1005 and center of mass determination software 1006 may be stored on the mass storage device 1004. One or more of the operating system 1005 and center of mass determination software 1006 (or some combination thereof) may comprise program modules and the center of mass determination software 1006. Sensor data 1007 may also be stored on the mass storage device 1004. Sensor data 1007 may be stored in any of one or more databases known in the art. The databases may be centralized or distributed across multiple locations within the network 1015.

[0072] A user may enter commands and information into the computing device 1001 using an input device (or plurality of input devices 1016, as illustrated in FIG. 18, comprising a

17 joystick and a touchscreen). Such input devices can comprise, but are not limited to, a joystick, a touchscreen display, a keyboard, a pointing device (e.g., a computer mouse, remote control), a microphone, a scanner, tactile input devices such as gloves, and other body coverings, motion sensor, speech recognition, and the like. These and other input devices may be connected to the one or more processors 1003 using a human machine interface 1002 that is coupled to the bus 1013, but may be connected by other interface and bus structures, such as a parallel port, game port, an IEEE 1394 Port (also known as a Firewire port), a serial port, network adapter 1008, and/or a universal serial bus (USB).

[0073] A display device 1011 may also be connected to the bus 1013 using an interface, such as a display adapter 1009. It is contemplated that the computing device 1001 may have more than one display adapter 1009 and the computing device 1001 may have more than one display device 1011. A display device 1011 may be a monitor, an LCD (Liquid Crystal Display), light emitting diode (LED) display, television, smart lens, smart glass, and / or a projector. In addition to the display device 1011, other output peripheral devices may comprise components such as speakers (not shown) and a printer (not shown) which may be connected to the computing device 1001 using Input/Output Interface 1010. Any step and/or result of the methods may be output (or caused to be output) in any form to an output device. Such output may be any form of visual representation, including, but not limited to, textual, graphical, animation, audio, tactile, and the like. The display 1011 and computing device 1001 may be part of one device, or separate devices.

[0074] The computing device 1001 may operate in a networked environment using logical connections to one or more remote computing devices 1014a,b,c. A remote computing device 1014a, b,c may be a personal computer, computing station (e.g., workstation), portable computer (e.g., laptop, mobile phone, tablet device), smart device (e.g., smartphone, smart watch, activity tracker, smart apparel, smart accessory), security and/or monitoring device, a server, a router, a network computer, a peer device, edge device or other common network node, and so on. Logical connections between the computing device 1001 and a remote computing device 1014a, b,c may be made using a network 1015, such as a local area network (LAN) and/or a general wide area network (WAN), or a Cloud-based network. Such network connections may be through a network adapter 1008. A network adapter 1008 may be implemented in both wired and wireless environments. Such networking environments are conventional and commonplace in dwellings, offices, enterprise-wide computer networks,

18 intranets, and the Internet. It is contemplated that the remote computing devices 1014a,b,c can optionally have some or all of the components disclosed as being part of computing device 1001. In various further aspects, it is contemplated that some or all aspects of data processing described herein can be performed via cloud computing on one or more servers or other remote computing devices. Accordingly, at least a portion of the system 1000 can be configured with internet connectivity.

[0075] Referring to FIG. 18, the user can control general movement of the wheelchair with the joystick, and the user can, via the touchscreen, toggle specific commands, such as raising up, lowering down, tilting forward, tilting backward, tilting left, tilting right, moving the carriage forward, or moving the carriage backward. The touch screen can further enable the user to execute more advanced features such as automated navigation and obstacle traversing.

Examples

Example 1

[0076] A. Electronics Design

[0077] The driving mechanism of an exemplary wheelchair comprises two drive wheels operated by ajoystick R-NET controller (Curtiss-Wright Corp., Cheswick, PA, USA), while the driving wheel carriage is controlled by two DC motors (Dumore Corp., Mauston, WI, USA) to adjust its drive-wheel configuration. Both drive wheel systems are equipped with incremental encoders (US-Digital, Vancouver, WA, USA) to monitor the drive wheels’ speed, acceleration, and translation with relation to the frame. Built-in position sensors measured the pneumatics’ movement, which were translated to wheel height. Additionally, a nine-degree-of-freedom inertial measurement unit (Adafruit, New York, NY, USA) was mounted under the seat to monitor stability of the exemplary wheelchair (pitch and roll angles). The multiple sensors were fed into an embedded system (Raspberry PI Foundation, Cambridge, U.K.) to control the height of each wheel and perform the control curb negotiation process.

[0078] B. Curb Negotiation Optimization

[0079] A semiautomatic curb negotiation application was created based upon electric powered wheelchair (EPW) user feedback. A solution was a collaboration between the user

19 and the exemplar _)' wheelchair, in which the device performs the curb negotiation automatically while the user remains in control of the process. Wheel height is controlled with a gain-scheduling controller using curb height as the desired position. The interface of the exemplary wheelchair, differing from a previous interface having slide potentiometers, included a five-switch keypad (SWi, i = 1,...,5) to activate any of the four advanced mobility applications: power seat functions (SWI), curb ascending mode (SW2), curb descending mode (SW3), and self-leveling mode (SW4). Once the curb ascending or descending mode was selected, SW5 (switch 5) was used to continue, pause, or reverse the application process (see FIG. 20). Accordingly, in exemplary aspects, switch 5 can be a three-position switch, having positions corresponding to forward, pause, and reverse.

[0080] Prior to ascending or descending a curb, the user was instructed to align the exemplary wheelchair perpendicular to the curb. This first step was completed by allowing the user to drive until both front wheels were in rolling contact with the curb when ascending or until both front wheels left the edge of the curb when descending. This was selected as a safety requirement to maintain the wheels in contact with the ground during the process. However, bench-test results demonstrated that the exemplar)' wheelchair was able to climb ~8.6° misalignment with the curb. Once the user perceives that the front wheels are in contact with or off the curb, the user activated the curb ascending or descending mode using SW2 or SW3, respectively.

[0081] In either mode, the user remained in control while holding SW5 forward. If SW5 was released, the exemplary wheelchair remained stationary until the user pushed SW5 forward again to continue the process or pushed SW5 backward to reverse the ascending process. This condition was added to allow users to pause and verify movements of the exemplary wheelchair during curb negotiation. During curb ascent, the exemplar)' wheelchair started the curb ascent by elevating the frame to the curb height while moving the wheel carriage backwards (see FIGS. llA-11C). For this study, the curb height was predetermined in the application algorithm to evaluate the reliability of the application in engineered curbs with known heights. This step shifted the weight toward the curb to ensure stability. This step was followed by lifting the front casters and driving wheels on top of the curb (see FIGS. 11D- 1 IF). The curb ascending process was completed when the wheels returned to their initial position of front-wheel drive.

20 [0082] During the curb descending mode, the front wheels were lowered automatically to the curb height, followed by driving off the curb and placing the driving wheels on the low surface (see FIGS. 12A-12C). Then, the exemplary wheelchair was configured in the rear- wheel-drive mode at maximum elevation to maintain stability, while the rear casters left the curb (see FIGS. 12D-12F). Once this step was completed, the exemplary wheelchair returned to the front-wheel-drive configuration and the wheels were moved to a lower-ground clearance.

[0083] III EVALUATION [0084] Participants

[0085] Inclusion criteria for participants were 18 years or older, weighed less than 113.4 kg (250 lb), had at least one year of experience using an EPW, and were able to tolerate sitting for 3 h. Participants with active pressure sores and/or back, pelvic, or thigh pain were excluded. The study was conducted at the National Veterans Wheelchair Games held in Orlando, Florida, and at the Human Engineering Research Laboratories in Pittsburgh, PA, USA. The study was approved by the Veterans Affairs Pittsburgh Healthcare System Institutional Review Boards (ID: PRO02495).

[0086] Protocol

[0087] Participants were first briefed and consented about the study and then screened to meet the eligibility criteria. After consent, participants completed a demographics questionnaire and information about their own EPWs. Then, participants received training on the exemplary wheelchair’s features and the curb negotiation application. During training, participants ascended and descended curbs a minimum of two times; the training continued until participants and researchers felt comfortable using the exemplary wheelchair. The training lasted approximately 15 min.

[0088] After training, participants performed driving tasks that simulated real-world environmental barriers encountered by EPW users. These driving tasks addressed environmental barriers, such as the absence of curb ramp (or curb cuts) or the presence of high curbs that do not comply with accessibility standards. The driving tasks included ascending and descending two curb heights of 152 mm (6.0 in) and 203 mm (8.0 in). Additionally, these curb heights were modeled based on a sidewalk height range by the

21 American Association of State Highway and Transportation Officials (AASHTO). These tasks were repeated for five trials. At the end of the study, participants were asked to complete the National Aeronautics and Space Administration-Total Load Index (NASA- TLX) and system usability scale (SUS) tools to evaluate their cognitive demand to perform the tasks and the usability of the device, respectively. Additionally, participants’ comments and recommendations were collected for further evaluation.

[0089] Outcome Measures

[0090] Effectiveness: Effectiveness was defined as how well the device accomplished the given tasks. Quantitative driving metrics, such as task completion time (seconds) and maximum/minimum changes in seat angle (pitch and roll angles), were collected to evaluate the exemplary wheelchair’s performance. These metrics were recorded to demonstrate design objectives D1 and D2.

[0091] Workload Demand and Usability Assessment Tools: Participants rated the NASA- TLX (HI) and SUS (H2) assessment tools to evaluate participants’ perception toward the exemplar wheelchair when negotiating curbs. The NASA-TLX tool measures the workload demand to perform a task in six subscales: mental, physical demand, performance, effort, and frustration. Participants rated each subscale on a scale of 0-100 and then selected the most valuable subscale between two. These scores were combined to obtain a weighted score for each subscale. A maximum overall NASA-TLX score of 50 was considered acceptable based on 1173 studies using the NASA-TLX as metric to measure workload demand.

[0092] The SUS questionnaire measures the usability of a device to perform tasks. The questionnaire includes ten questions rated at a five-point Likert scale, 1 — “Strongly Disagree” to 5 — “Strongly Agree”. The overall score ranges from 0 to 100, in which a SUS score of 69 is considered an acceptable score of usability among different studies. Each assessment tool included a comment section in each subscale, in which participants could describe their responses toward the exemplary wheelchair and its curb negotiation application.

[0093] 3) Data Analysis: Demographics and categorical metrics, such as completion time and angle deviations, were described in means and standard deviation (SD), whereas perception metrics, such as the NASA-TLX and SUS subscores, were described in interquartile (IQR 1- 3), medians, and box plots. The average and SD completion time to ascend and descend curbs

22 were combined to meet Dl. In addition, average and SD maximum/minimum seating angle deviations were recorded to meet D2. The seat angle changes at the beginning of the curb ascending process were not considered for post-processing analysis because the seat was tilted intentionally to place the front wheels on top of the curb height. The effectiveness of the curb negotiation application was evaluated in different curb conditions; the curb ascending/descending completion time and seat angle changes between two curb heights were compared using a pairwise t-test. The driving tasks were labeled in inches for good visualization of the results: CA6 — Curb Ascending 6 in (152 mm), CD6 — Curb Descending 6 in, CA8 — Curb Ascending 8 in (200 mm), and CD8 — Curb Descending 8 in.

[0094] To test HI and H2, the NASA-TLX and SUS overall scores were compared to their acceptable score references. The subscales of each assessment tool were analyzed individually. All statistical analyses were performed using Statistical Package for the Social Science version 25.0 (SPSS Inc., Chicago, IL, USA). The level of significance was set at a = 0.05 for all comparisons.

[0095] IV. RESULTS

[0096] Demographics

[0097] Ten experienced EPW users evaluated the exemplary wheelchair. Expenenced EPW users drove the exemplary wheelchair over two curb heights for five trials at each height and rated the exemplary wheelchair in terms of effectiveness, workload demand, and usability. Participants included eight males and two females with an average age of 59.3 +/- 12.6 years, an average mass of 91.8 +/- 10.7kg, and an average EPW driving experience of 11.6 +/- 6.9 years. Participants reported multiple mobility impairments including spinal cord injury and multiple sclerosis. Seven participants used a mid-wheel-drive, two front-wheel-drive, and one rear-wheel-drive EPW.

[0098] Effectiveness

[0099] Eight participants successfully ascended and descended two curbs of 152 mm (6 in) and 203 mm (8 in) heights for five trials, while two participants stopped after three trials of each task. These two participants stated that the protocol was repetitive and their perception toward the curb negotiation application was not going to be affected by performing more trials.

23 [0100] The exemplary wheelchair required 51.6 +/- 18.1 and 59.8 +/- 20.2 s to complete CA6 and CA8, respectively. In addition, CD6 was completed in 29.4 +/- 7.7 s compared to 31.3 +/- 10.3 s required to complete CD8. These results showed a significant time difference, which was likely associated with less range of motion required for the exemplary wheelchair to reach the curb height (see Table I). The combined average time to ascend and descend a 152-mm curb was 81.0 +/- 24.7 s while completing a 203-mm curb required 91.1 +/- 26.4 s. These results showed a time improvement compared to the previous curb negotiation application; however, it was longer than the street signalization cycle length of 30-60 s (Dl). While the automation of the curb negotiation process and interface modifications improved the completion time; mechanical changes will be proposed in Section V for further improvement.

[0101] The seat angle changes at the beginning of the curb ascending process were not considered for postprocessing analysis because the seat was tilted intentionally to place the front wheels on top of the curb (see FIG. 11 A). The average minimum and maximum pitch angles for each task were within the safety angle threshold of <4.5° w.r.t. a horizontal ground (D2), as shown in Table I. Likewise, the roll angle in each task, except for the maximum roll angle in CD8, was within the angle threshold. During the curb descending process, the driving wheels dropped off the curb in FIG. 12C, which shifted the center of mass toward the front wheels. This caused the pneumatics of the front wheels to compress due to their lifting limitations and continuously adjust the seat angles until the wheels reached the desired position. This behavior was observed at the end of the curb descending process (see FIGS. 12E and 12F) when all the wheels moved to their initial position. While it is important to address this behavior, a previous study demonstrated the exemplary wheelchair’s stability to maintain the center of mass within its footprint at greater seat angles.

24 TABLE I: MEAN AND SD OF EACH TASK COMPLETION TIME (SECONDS) AND MAX/MIN CHANGES IN THE SEAT ANGLES FOR EACH TASK

Tasks were labeled as curb ascending (CA) and curb descending (CD) for 6.0- and 8.0-in curb heights. *P-value is significant at <0.05. «Results did not meet D2

[0102] Task Load Demand and Usability Measures

[0103] Participants rated the ease of use of the exemplary wheelchair to negotiate curbs and their cognitive demand to operate the exemplary wheelchair (see Table II). Seven participants agreed that the exemplary wheelchair’s curb negotiation application was quick to leam and easy to use. Furthermore, eight out of ten participants felt confident in operating the exemplary wheelchair over curbs of different heights. These results were complemented by the NASA-TLX subscores, in which participant rated low cognitive (mental, physical, temporal, and frustration) demand with confidence in performing the assigned tasks. These results reflected the usability of the interface and level of autonomy of the application compared to its previous iteration.

[0104] V. DISCUSSION

[0105] A. Effectiveness

[0106] The exemplary wheelchair successfully ascended and descended curbs of different heights, and the completion time was significantly improved compared to its previous

25 iteration. However, D1 was not supported by the results. Three participants considered the speed of the curb negotiation process to be comfortable, while others highlighted that the process should be faster, particularly when driving outside and in weather conditions where the user may be at risk if exposed for a long period of time.

TABLE II: PARTICIPANT’S PERCEPTION TOWARD CURB NEGOTIATION APPLICATION IN TERMS OF USABILITY (SUS) AND TASK LOAD DEMAND

(NASA-TLX)

* Overall scores meet HI andH2.

[0107] It was observed that the wheels required time to elevate or lower to the curb height throughout every ascending and descending step (ap-proximately 5-10 s per step). Four participants were concerned about the ‘ ^“ bounciness” of the device, in which the wheelchair continuously adjusted until the wheels reached the desired height. This effect was caused by the pneumatics’ lifting limitations and its nonlinear response to control due to several factors (i.e., user’s weight, air temperature, air volume, and air dissipation). In addition, one participant was more concerned about the number of times the exemplary wheelchair was able to negotiate a curb before the air supply would limit the exemplary wheelchair’s ability to ascend and descend curbs. Alternative actuators, such as electrohydraulic (E-H) actuators, are recommended due to fast response time and linear control. E-H actuators are popular in quadruped legged robots that traverse unknown environments due to their lifting capabilities, ease of control, and fast movement. These actuators provide similar lifting capabilities to the

26 pneumatic actuators used in the current prototype. E-H actuators (Kaman Fluid, Pittsburgh, PA, USA) are powered by batteries, which eliminates the need for an air supply.

[0108] The driving performance of the exemplary wheelchair application was evaluated for efficiency in different curb conditions, which showed a completion time difference between descending and ascending curbs of different heights. A reason was the time required to adjust the wheels to the curb height; hence, a smaller curb height required less adjustment time. The curb ascending mode required twice the time as the curb descending mode due to the extra sequential steps to move the drive wheel carriage.

[0109] While the combined completion time to ascend and descend a curb did not meet Dl, the particular ability to either ascend or descend a curb of different heights is important to access places of public interest (e.g., church, restaurant, and store), where a single curb might be present. Commercial EPWs provide curb ascending capabilities up to 50 mm (2 in). However, their ascending strategies may entail driving straight toward the curb at slow speeds for shallow heights or a combination of full speed and shift in the center of mass for higher curb heights. These strategies may put the user at risk of tipping and compromise the lifespan of the EPW. Despite the time that the exemplary wheelchair requires to negotiate curbs, it enhances accessibility in a safe manner for a wide range of curb heights. Nine out of ten participants were neutral to or agreed that they would use the application frequently, addressing the importance of negotiating curbs to access locations with high structures.

[0110] High changes in the seat pitch and roll angles were observed at steps (see FIG. 12B and FIG. 12C) of the descending process, which was caused by an algorithm limitation in step (a). When the descending process was activated, the front casters were extended to the length of the curb height. However, the software did not take into consideration an additional ground clearance gap of 1.5 in below the wheelchair frame. This gap inhibited the front casters from touching the ground, which caused a forward dip when moving forward in steps (b) and (c). Regardless of the seat changes, the exemplary wheelchair maintained stability during these steps. Further exemplar wheelchairs can incorporate force gauges to provide additional feedback when the wheels are in contact with the ground during the curb negotiation process.

[0111] B. Subjective Measures

27 [0112] Participants indicated an acceptable NASA-TLX overall score when operating the exemplar wheelchair with the semiautonomous CA/CD application. These scores complemented the ease of use and intuitiveness of the interface (see Table II) compared to its previous iteration that allowed users to operate each wheel with slide potentiometers. On the other hand, participants reported mixed feelings about the level of automation of the application. Participants stated preference for a single button to perform an automated application, while others preferred to be in control of the wheelchair during the curb negotiation process.

[0113] A common theme by all participants was the need for visual feedback to monitor the exemplar wheelchair’s movements for safety and environment awareness. Furthermore, it is important to address fault tolerance when EPW users request the incorrect applications. Participants received training and successfully completed the trials independently; however, visual feed-back (e.g., changes in LED colors and graphical user interface (GUI)) is needed to confirm wheelchair user’s action as addressed by participants. The research offers a sensory system using Intel RealSense (Santa Clara, CA, USA) cameras that identifies curb characteristics for path planning prior ascending or descending curbs. Additionally, the sensory system offers a GUI to display the curb height or curb drop when the user requests to ascend or descend the curb, respectively. One participant recommended the use of padding in the switches because they were uncomfortable. All participants showed dexterity to operate the interface, people with less upper range of motion may require alternative interfaces, which is possible with the exemplary wheelchair. The use of a GUI reiterates the need of customizable options to select different applications.

[0114] Participants reported “neutral to slightly agree” on their frequency to use the application. Their responses might vary due to their level of accessibility in participants’ communities. The exemplary wheelchair allows people to access points of interests that they are unable to negotiate due to environmental barriers and current EPW limitations. In terms of study limitations, participants’ scores might have been influenced by the presence of researchers and other participants (Hawthorne effect). To reduce this bias, researchers performed the same protocol and engaged the same way with each participant. Second, the study assumes ideal conditions; therefore, further evaluation is necessary to consider other factors, such as changeable environments, weather conditions, and sloped curbs, which affect the driving performance of robotic wheelchairs. The sample size was limited to ten

28 participants due to the limited population of EPW users where the study was conducted. It has been suggested that 80% of usability feedback can be detected with four or five participants.

[0115] The proposed study highlighted the need to ascend and descend curbs standardized by the AASHTO that commercial EPWs are unable to climb and accessibility to points of interests where ADA guidelines are not met. The results of this study provide novel evidence of usability of robotic wheelchairs for obstacle negotiation with end-users using standardized quantitative and qualitative metrics. The study contributes to the literature by outlining requirements of robotic wheelchairs to address user needs when facing less accessible environments.

Example 2

[0116] An exemplary w'heelchair comprises six independently height-adjustable wheels with a modular drive-wheel configuration, omni-wheels as caster wheels to eliminate swivel, and a footprint comparable to that of commercially available EPWs. Each wheel was linked to an active suspension (AS) system that included an adjustable pneumatic shock absorber and an electro-hydraulic motor in series (FIG. 14A). Shock absorbers provided a passive suspension to reduce vibration on uneven surfaces similar to EPWs; while electro-hydraulics were automatically controlled to maintain stability when surface irregularities (e.g., inclined surfaces) were detected. It was contemplated that electro-hydraulics can reduce whole body vibrations (WBV) in conjunction with shock absorbers when driving on surfaces transitions that combine an inclined surface with a threshold. This study compared the exemplary wheelchair with active suspension (AS), the exemplary wheelchair without active suspension (no AS), and a commercial EPW with passive suspension.

[0117] No AS refers to the exemplary wheelchair with inhibited electro-hydraulic actuators and is only reliant on its shock absorbers. The shock absorbers were air-pressured, adjustable, and set at 100 psi per wheel. The selected commercial EPW was the Permobil F5 Corpus, a front-wheel-drive EPW with shock absorbers (FIG. 14B) in each wheel to ameliorate WBV exposure. EPWs with a front-wheel-drive configuration assist with obstacle climbing, stability, and traction outdoors. Both EPWs used an R-Net controller to configure the same driving parameters (i.e., speed and acceleration).

29 [0118] The Shimmer 3 triaxial accelerometer (Shimmer, Boston, MA, USA) was mounted in the seat pan of each EPW with its z+ axis facing orthogonal to the seat. The accelerometer incorporates a stand-alone microcontroller (STMicro LSM303AHTR) with a 14-bit resolution, high sensitivity (to detect +/ -8 g), and at a sampling frequency of 100 Hz. The sampling frequency was selected in order to identify a suitable range of frequencies between 0.01 and 80 Hz according to the ISO 2631:1 standard. Similar studies acquired vibration data through accelerometers at a sampling frequency between 50 and 102 Hz. The sensor was validated for use in human health monitoring, monitoring activities of daily living, and environmental and habitat monitoring.

[0119] Its accelerations were used to calculate the root mean square (RMS) and vibration dose values (VDV) following the ISO 2631:1 (1997) standard. A rehabilitation seating cushion was used following the ISO 16840-2 wheelchair seating standards. The cushion material was made of high-density foam developed to support bariatric loads and to manage tissue integrity. The cushion was previously used in Garcia's and DiGiovine's studies, which showed a transmissibility coefficient of 1.2 and 0.5, respectively. The cushion transmissibility was approximated to 1; therefore, the seat pan and cushion showed similar WBV exposures. Additionally, the accelerometer was placed under the seat cushion to prevent it from moving during testing and to measure vibrations directly from the rigid body.

[0120] 2.2. Protocol

[0121] A 50th percentile Hybrid II anthropometric dummy of 100 kg was used to simulate a person seated in each EPW. Three trials were performed by driving each EPW on five selected surfaces for a total of 45 trials. Each EPW was controlled remotely by a researcher. The wheelchair speed was set to 1.2 m/s, which is the same as an average person's speed when walking across the street 127] A MATLAB Graphical User Interface (GUI) was developed to measure the time-series accelerations in real-time during the completion of each trial. The GUI facilitated data collection by connecting to the accelerometer, recording data, and saving the data to a custom filename.

[0122] 2.3. Surfaces

[0123] Five engineered driving surfaces were selected to represent surface transitions that EPW users commonly encounter daily. The tasks included: going up and down a 10° slope with and without a 2.5 cm threshold (FIG. 15A,B), surfaces with a roughness of 12.5 cm/m

30 and 18.3 cm/m (FIG. 15C), and a series of potholes of up to 30.5 cm in diameter and 5.0 cm in depth (FIG. 15D). The 10° slope simulated conventional incline and decline ramps considered worst-case scenarios for wheelchair dynamic stability as part of the ANSI/RESNA wheelchair standards ISO 7176-2.

[0124] FIG. 15A shows an Up-Flat-Down 10° Ramp, FIG. 15B shows an Up-Flat-Down 10° Ramp with a 2.5 cm threshold in transition, FIG. 15C shows Surface roughness with adaptable slabs, and FIG. 15D shows Potholes of 5.4 cm in depth.

[0125] The slopes measured 3.1 m in length and 1.2 m in width. The 2.5 cm surface threshold simulated non- ADA thresholds obtained from Bennef s study that reported a curb-ramps threshold of 1.9 +/- 0.1 cm. The ADAAG guidelines recommend a maximum 0.6 cm threshold in lip height for water drainage. Two surfaces of 12.5 cm/m and 18.3 cm/m in roughness represented uneven sidewalks and rough terrains. Both tasks were 1.2 m wide by 2.4 m long and used wooden slabs of 1.9 cm in height. Last, a series of potholes were simulated based on Kirby's wheelchair skills test v.4.1 that included 5.0 cm deep potholes across a 2.4 m long by 1.2 m wide surface. The surface represented potholes caused by wear- and-tear due to weather conditions and constant use by heavy -load vehicles to which wheelchair users are exposed.

[0126] 2.4. Data Analysis

[0127] Descriptive analysis (e.g., means, standard deviation) and bar graphs described the WBV parameters of each EPW in terms of RMS and VDV. These variables were calculated using the raw acceleration data obtained from the triaxial accelerometer. The raw acceleration data (ax, ay, az) were first multiplied by their frequency weighting in terms of comfort (kx = ky = kz = 1). The study focused on vibration values in the seat pan because the back rest and footplate were fixed to the seat and the seat was considered as a rigid body. The weighted accelerations were then calibrated and processed for analysis using MATLAB. The first and last second time-stamp of the data were cut-off to eliminate zero values. Then, the raw accelerations in three axes were calibrated by subtracting the mean of each axis from its corresponding value. The total magnitude of each acceleration w as calculated and used in the data analysis for this study. The total acceleration was viable to use because WBVs have an effect in multiple directions. Previous studies were limited in only looking at accelerations in the vertical (or gravity) direction; however, this limits the analysis to only vertical

31 accelerations and not lateral and front/back accelerations. The RMS and VDV were calculated using the total calibrated acceleration with the equations below:

(1) a _Totd = (kx^x + ky y i k i) where T is the duration of the trials.

[0128] Statistical analyses were performed using the IBM SPSS software version 24.0 (SPSS, Inc., Chicago, IL, USA). Data were analyzed for normality using the Shapiro Wilk test. One-way ANOVA was performed to compare the RMS and VDV mean differences between surfaces transitions within each EPW to evaluate the effects of a surface threshold/gradient. The same statistical test was performed to compare the RMS and VDVD mean differences between EPWs for each surface transition to evaluate EPWs' suspensions. The level of significance was set at cc = 0.05 for all comparisons. If results were significant, post-hoc analysis was performed with a Bonferroni correction to adjust for multiple comparisons.

[0129] 3. Results

[0130] Results show no significant differences in average WBV (RMS and VDV) values between the commercial EPW, the exemplary wheelchair with AS, and the exemplary wheelchair with no AS when driving on surface transitions with different thresholds (FIG. 16).

[0131] In terms of WBV differences between surfaces, it was found that potholes caused significantly higher RMS values on the commercial EPW compared with the 12.5 cm/m surface roughness (p < 0.001) and the 10° ramp without threshold (p < 0.001) (Table III). Additionally, the 10° ramp with a threshold showed higher RMS values on the commercial EPW compared with the 12.5 cm/m surface roughness (p = 0.002) and the 10° ramp without a threshold (p = 0.002). However, there were no significant RMS differences between surfaces w hen using the exemplary wheelchair with or with no AS (FIG. 17).

32 Table III. Root-mean-square (RMS) and vibration dose value (VDV) of tested devices per surface

[0132] The VDV differences between surfaces in each EPW were more noticeable with an increase in the surface threshold. The commercial EPW and the exemplary wheelchair with AS showed significantly lower VDV values when driving on the 10° ramp without a threshold compared with potholes of 5.0 cm in depth (p < 0.001). Likewise, the Permobil F5 and the exemplar}' wheelchair with AS showed significantly higher VDV exposure when driving on the 10° ramp with a 2.5 cm threshold compared with no threshold (p < 0.001). Additionally, the surface with potholes reported significantly higher VDV values compared with surface roughness with a 1.9 cm threshold (p < 0.001). No significant WBV differences were found when driving the exemplary wheelchair with no AS across all surfaces.

33 [0133] FIG. 17 illustrates RMS total acceleration (Top) and VDV (Bottom) differences between surfaces with each EPW. Significant differences between surfaces are denoted with an asterisk (* p-value < 0.01 post-hoc Bonferrom correction).

[0134] 4. Discussion

[0135] 4.1. WBV Differences between Surfaces

[0136] The results showed that surfaces with a threshold over 2.5 cm showed high RMS values over 1.2 m/s ² and a maximum RMS of 1.7 m/s ² when driving a commercial EPW. The ISO 2631-1 standard suggests that a RMS acceleration of 1.6 m/s ² or greater could be harmful over a 1-h period. This is approximately the daily duration that wheelchair users drive their assistive devices each day. This finding suggests that the passive suspension of commercial EPWs and the exemplary wheelchair is beneficial to reduce vibration exposure on surfaces with different thresholds.

[0137] It is worth noting that the Permobil F5 is a high-end EPW with a front- wheel-drive configuration and all-terrain wheels designed to traverse environments with a threshold of up to 3.0 in. in height, according to the manufacturer. The availability of high-end EPWs as such depends on the user's level of impairment and insurance coverage. Alternative cost-effective EPWs have less weight for easier transportation but are limited to fewer seating features and less efficient drive motors. Additionally, the suspension dampening required to ameliorate WBV effects, particularly on these surfaces, is unknown. Further evaluation of WBV effects on EPWs is encouraged to reduce EPW users' discomfort on surfaces with thresholds.

[0138] The exemplary wheelchair’s EPW with active suspension was introduced in this study as an alternative suspension mechanism that combines a shock absorber and an electro- hydraulic actuator in series. There were no significant RMS differences between surface thresholds whether using the exemplary wheelchair with or without the AS mechanism. The vibration for each surface remained below 1.2 m/s ² expect for potholes and the 10° ramp with a 2.5 cm threshold. These results demonstrated that the exemplary wheelchair can reduce the vibration with the use of shock absorbers alone; on the other hand, the use of actuators in the exemplar wheelchair for active suspension remains important to maintain stability on inclined and uneven surfaces to reduce tips and falls.

34 [0139] The findings showed that VDV values increased with the surface threshold. For instance, the surface with potholes of 5.4 cm in depth showed significantly higher VDV values than the 10° ramp (with no threshold) and surface roughness with a 1.9 cm threshold. These results are consistent with those of the Permobil F5 and the exemplary wheelchair with AS. The VDV is more sensitive to the acceleration peaks; therefore, it is contemplated that the amount of vibration for the case of inherent shock exposure can be estimated. Further, these results are consistent with a study that evaluated an EPW at a speed range of 0-0.8 m/s on a 3.6 cm threshold and showed a VDV range of 0.7-2.25 m/s ¹ · ⁷⁵ , respectively. While the EPW speed was constant at 1.5 m/s, it is expected that faster speeds can increase the likelihood of higher VDV exposure. EPW users tend to avoid surfaces with high thresholds to avoid discomfort and reduce the risk of tipping or falling; however, these surfaces may be inevitable when alternative routes are not accessible nor available. Further studies should look into automatically changing the speed of the EPW on high WBV surfaces to reduce RMS and VDV values.

[0140] Further, the 10° ramp with a 2.5 cm threshold showed high VDV values compared with no threshold (FIG. 17). Although the results were below the health vibration zone of 9.0 m/s ¹ · ⁷⁵ , a high number of curb-ramps do not meet the ADA standards of a maximum threshold of 0.6 cm usually found in real-world conditions for water drainage. Driving on these thresholds exposes EPW users to muscle pain and discomfort and can damage the assistive device over time. Damaging EPW users' only source of assistive mobility limits their participation in the community and renders them unable to perform leisure and vocational activities. It is contemplated that adequate settings in the EPW such as shock absorber dampening, drive wheel suspension, and cushioning can reduce discomfort and WBV exposure when environmental barriers are present. Additional mechanisms that can help reduce WBV not currently in EPWs are contemplated.

[0141] 4.2. WBV Differences between EPWs

[0142] The results show no significant RMS and VDV differences between EPWs' suspensions across all surface thresholds, and these were below the health caution zone of 1.6 m/s ² over 1 h of exposure at the comfort level. EPWs serve as a means of mobility for users to commute from home to work/school/shops, particularly when public or private transportation is not available or accessible. The typical EPW user can drive at least 1 h/day assuming a normal speed of 1 m/s between locations. During travel, EPW users may drive on

35 sidewalks and roads with a threshold of over 2.5 cm in height. Further, EPW users are more exposed to these surface thresholds on sidewalk elevations due to tree growth on paved sidewalks and a lack of maintenance. While WBV values remained below the health risk threshold, these values can increase with additional elements within the EPW such as speed, cushioning, longevity, and weight.

[0143] FIG. 16 shows high WBV variance in the exemplary wheelchair with AS and no AS across all surfaces. A possible cause is the low dampening settings of the shock absorbers, which caused a high degree of displacement of its suspension. Likewise, a delay in the activation of the legged-wheel actuators in the AS system may have caused the EPW to replicate the surface profile, causing a bounce effect. On the other hand, the WBV variance in the EPW was only noticeable when driving on the surface with 5.0 cm potholes.

Additionally, the crash dummy also plays a passive role compared with a real end-user who may intentionally correct his/her posture and, hence, reducing the WBV variance.

[0144] The active suspension of the exemplary wheelchair did not reduce nor increase the vibration effects when traversing surface thresholds. The exemplary wheelchair’s active suspension was designed to prevent tips and falls when driving on inclined surfaces by adjusting its legged-wheel actuators in the base. Likewise, the goal of the shock absorbers was to serve as a form of passive suspension to reduce vibration. The results only demonstrated that its actuators can be inhibited when driving on surfaces with thresholds to improve power consumption while prioritizing shock absorbers in these surface conditions.

[0145] 4.3. Limitations

[0146] The study was conducted within lab settings and in controlled environments. Real- world surfaces may be affected by wear-and-tear due to weather conditions and pedestrian/vehicle traffic that we may not have included. Additionally, the wheelchair and user might be exposed to other sources of vibrations, such as large vehicles and construction sites.

[0147] A constant speed was used, which might not be typical when facing these types of surfaces. In addition to surface thresholds, driving at a faster speed may induce higher WBV magnitudes. EPW users tend to slow down when facing irregular and unfamiliar surfaces. These factors can be observed in further studies on WBV exposure in EPWs and other mobility assistive devices.

36 [0148] The commercial EPW used had a front- wheel-drive configuration that is mostly used for active wheelchair users in the community. However, there are other drive wheel configurations (mid- and rear- wheel drive) where WBVs may have different effects on EPW users. For example, mid-wheel drive provides high stability on flat surfaces and a small turning radius, but it is at risk of getting stuck on small thresholds and ramps. Rear- wheel drive is a less common in EPWs and mostly used outdoors due to its fast speed but it is prone to tipping as its center of mass is located towards the back and its front wheels may be smaller than the suggested ADA thresholds.

[0149] Finally, EPW users were not recruited for the study to avoid WBV exposure. A crash test dummy was used to prevent discomfort to end-users and for safety when operating the EPWs on challenging surfaces. Additionally, using a crash test dummy provided control over other factors that may influence the vibration exposure, such as weight shifting, repositioning, and weight distribution, commonly encountered with end-users. On the other hand, EPW users can provide feedback in terms of health and comfort when exposed to the vibration levels on surface transitions. Their feedback is important to be able to offer the most adequate mobility assistive device to reduce WBV exposure. Further studies may include subject testing to evaluate the feasibility of EPW suspensions.

[0150] 5. Conclusions

[0151] This study aimed to explore the WBV effects in EPWs when encountering challenging surfaces. While many studies have evaluated vibration in manual wheelchairs, there are few studies that evaluate the vibration effects in EPWs, particularly on surfaces with thresholds that end-users are exposed to daily. This is the first study to evaluate EPW suspensions on surfaces with different thresholds (heights) such as uneven sidewalks and curb-ramps that are not ADA-compliant. Likewise, this is the first study to compare two types of EPW suspension systems (passive and active suspension) to reduce WBV measures on the selected surfaces. The study introduced a novel EPW with active suspension to increase stability and the user's comfort. The results show similar WBV values that he within the health guidance safet zone; therefore, no difference was found between passive and active EPW suspension. The study also demonstrated a proportional increase in RMS and VDV values with the surface threshold when using the EPW passive suspension compared with the EPW active suspension, which demonstrated constant vibration values in all surface thresholds. The results of this study will increase the amount on literature of WBV exposure

37 in EPWs. This study was also conducted with a crash dummy and in controlled environments for safety reasons. Further evaluation of EPW suspension systems should include end-users to obtain their perception of comfort and health with respect to the WBV exposure in every day environments.

EXEMPLARY ASPECTS

[0152] In view of the descnbed products, systems, and methods and variations thereof, herein below are described certain more particularly described aspects of the invention. These particularly recited aspects should not however be interpreted to have any limiting effect on any different claims containing different or more general teachings described herein, or that the ‘ ^“ particular” aspects are somehow limited in some way other than the inherent meanings of the language literally used therein.

[0153] Aspect 1: A wheelchair comprising: a seat; a frame coupled to the seat; a plurality of wheel assemblies, each wheel assembly comprising: a swing arm that is pivotably coupled to the frame; a wheel rotatably coupled to the swing arm about a respective rotational axis; an actuator system that is configured to adjust the position of the wheel relative to the frame, wherein the actuator system comprises: an actively controlled actuator; and a passive shock absorber arranged in series with the actively controlled actuator; a force sensor that is configured to provide an output associated with a ground force applied by the ground to the wheel; a position sensor that is configured to provide an output associated with a position of the wheel; a memory storing therein at least one center of mass threshold; at least one processor in communication with the memory and in communication with the force sensor and the position sensor of each wheel assembly, wherein the at least one processor is configured to:

38 determine, based on the outputs of the force sensor and position sensor of each wheel assembly of the plurality of wheel assemblies, a center of mass of the wheelchair; compare the center of mass of the wheelchair to the at least one center of mass threshold; and adjust, using the actively controlled actuator of at least one wheel assembly of the plurality of wheel assemblies, the center of mass of the wheelchair if the center of mass exceeds the at least one threshold.

[0154] Aspect 2: The wheelchair of aspect 1, wherein the at least one processor is configured to inhibit the actuator of each wheel assembly of the plurality of wheel assemblies from actuating.

[0155] Aspect 3: The wheelchair of aspect 1 or aspect 2, wherein the at least one center of mass threshold comprises a forward seat pitch, a rearward seat pitch, a left seat roll, a right seat roll, or combinations thereof.

[0156] Aspect 4: The wheelchair of any one of aspects 1-3, wherein at least one wheel assembly of the plurality of w heel assemblies is a caster wheel assembly, wherein the wheel of the caster wheel assembly is freely rotatable about the respective rotational axis.

[0157] Aspect 5: The wheelchair of aspect 4, wherein the wheel of the caster wheel assembly is coupled to the swing arm via a torsion joint, wherein the torsion joint is configured to enable pivotal movement of the wheel of the caster wheel assembly relative to a longitudinal axis that extends between a front and a rear of the wheelchair.

[0158] Aspect 6: The wheelchair of aspect 4 or aspect 5, wherein only one wheel assembly of the plurality of wheel assemblies is a caster wheel assembly.

[0159] Aspect 7: The wheelchair of any one of the preceding aspects, wherein the center of mass of the wheelchair accounts for a weight of a user in the seat of the wheelchair.

[0160] Aspect 8: The wheelchair of any one of the preceding aspects, wherein the actively controlled actuator of the actuator system of at least one wheel assembly of the plurality of wheel assemblies comprises a piston rod that is movable in a first direction, wherein movement of the piston rod in the first direction causes movement of the passive shock

39 absorber of said actuator system of said at least one wheel assembly in a second direction that is opposite the first direction.

[0161] Aspect 9: The wheelchair of any one of the preceding aspects, wherein the passive shock absorber is a linear shock absorber actuator having a first end and a second end that is axially movable relative to the first end.

[0162] Aspect 10: The wheelchair of any one of the preceding aspects, wherein the actively controlled actuator comprises an electrohydraulic cylinder and a piston rod that moves axially along the electrohydraulic cylinder.

[0163] Aspect 11 : The wheelchair of aspect 10, wherein the piston rod that moves axially along the electrohydraulic cylinder along a first axis, wherein the passive shock absorber is a linear shock absorber actuator having a first end and a second end that is axially movable relative to the first end along a second axis, wherein the first axis is within 15 degrees of parallel to the second axis.

[0164] Aspect 12: The wheelchair of any one of the preceding aspects, wherein the position sensor of each wheel assembly of the plurality of wheel assemblies is a rotational position sensor that is configured to provide an output based on a rotational position of the respective swing arm of the wheel assembly of the plurality of wheel assemblies.

[0165] Aspect 13: The wheelchair of any one of the preceding aspects, further comprising an acceleration sensor that is configured to detect vibration amplitude, wherein the acceleration sensor is in communication with the at least one processor, wherein the at least one processor is configured to cease adjustment of the center of mass of the wheelchair if the center of mass is within the threshold and the vibration amplitude is above a vibration threshold.

[0166] Aspect 14: The wheelchair of any one of the preceding aspects, further comprising a machine vision sensor in communication with the at least one processor, wherein the at least one processor is configured to modify a path of the wheelchair based on the machine visions sensor detecting an object.

[0167] Aspect 15: The wheelchair of any one of the preceding aspects, further comprising a speed sensor in communication with the at least one processor, wherein the processor is

40 configured to regulate a maximum speed of the wheelchair based on a proximity of the center of mass to the at least one threshold.

[0168] Aspect 16: The wheelchair of any one of the preceding aspects, wherein the passive shock absorber of the actuator system of at least one wheel assembly of the plurality of wheel assemblies is configured to adjust a dampening impedance based on a proximity of the center of mass to the at least one threshold.

[0169] Aspect 17: The wheelchair of any one of the preceding aspects, wherein the passive shock absorber of the actuator system of at least one wheel assembly of the plurality of wheel assemblies is configured lock based on a proximity of the center of mass to the at least one threshold.

[0170] Aspect 18: The wheelchair of any one of the preceding aspects, wherein the at least one processor is configured to determine a future center of mass based on a current center of mass of the wheelchair and an angular trajectory of the frame.

[0171] Aspect 19: The wheelchair of any one of the preceding aspects, further comprising an output device, wherein the at least one processor is configured to determine a weight of a user in the wheelchair, and cause the output device to output the weight of the user.

[0172] Aspect 20: A wheelchair comprising: a seat; a frame coupled to the seat; a plurality of wheel assemblies, each wheel assembly comprising: a swing arm that is pivotably coupled to the frame; a wheel rotatably coupled to the swing arm about a respective rotational axis; an actuator system that is configured to adjust the position of the wheel relative to the frame, wherein the actuator system comprises: an actively controlled actuator; and a passive shock absorber arranged in series with the actively controlled actuator.

[0173] Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the method and

41 compositions described herein. Such equivalents are intended to be encompassed by the following claims.

42