FIEED OF THE INVENTION:

The present invention relates to voice assisted devices and more particularly relates to a voice assisted device for object updating and query processing in the civil engineering domain based on machine learning with Artificial Intelligence (Al).

BACKGROUND OF THE INVENTION:

The field of civil engineering hinges upon the precise integration of client specifications, regional standards, and local codes governing engineering practices. In contemporary civil engineering workflows, these essential guidelines and requirements are conventionally delivered in electronic format at the outset of a project. Furthermore, comprehensive repositories of relevant standards and codes, such as the Bureau of Indian Standards (BIS) in India and the American Society of Civil Engineers (ASCE) and the American Concrete Institute (ACI) in the United States, are made readily accessible via their official websites.

These standards and codes are pivotal in ensuring the structural integrity, safety, and compliance of civil engineering projects. They encompass a wide spectrum of factors including design principles, material specifications, construction methodologies, and environmental considerations. However, the effective navigation and interpretation of these intricate standards and codes often pose a formidable challenge, particularly for those newly introduced to the multifaceted domain of civil engineering.

Civil engineering codes hold profound significance in the realm of structural safety and infrastructure reliability. These codes serve as comprehensive compendiums of guidelines and stipulations, governing diverse facets of construction, encompassing elements like architectural design, materials selection, construction methodologies, and the incorporation of environmental considerations. Yet, the intricate landscape of these codes and standards poses a formidable challenge, especially for individuals venturing into the multifaceted domain of civil engineering.

In the orchestration of construction projects, these specifications and standards assume a pivotal role, as they form the bedrock for project approval and serve as the legal foundation upon which projects are built. However, it is imperative to acknowledge that these standards and codes are subject to regular updates. The usage of outdated information can have detrimental repercussions on project approval and may lead to construction delays. The interpretation of these codes, standards, and specifications typically falls within the purview of seasoned engineering experts, a resource that is increasingly scarce. It generally demands a decade or more of practical experience for engineers to attain the level of proficiency required for these complex standards and specifications.

Failure to adhere to specifications or code provisions can precipitate project delays, often necessitating additional work for rectification or modification of project documents to align with the prescribed standards and specifications. These alterations incur supplementary costs, burdening the overall project. Given that standards and codes undergo periodic updates, the inadvertent omission of critical clauses can lead to substantial commercial penalties.

The prior art has witnessed the emergence of various software applications designed to extract essential data from codes and assess their interdependencies in the domain of civil engineering. These applications serve as valuable tools, enabling users to swiftly and accurately retrieve vital information from codes. By doing so, they streamline processes, saving time and mitigating the risk of costly errors. Such data extraction applications are available in diverse formats, including software, mobile applications, and web-based tools, ensuring widespread accessibility for civil engineers worldwide.

However, it is pertinent to recognize that the applications featured in the prior art exhibit limitations. They lack interactivity and fall short in ensuring precision. Furthermore, these applications place the onus on users to manually input search queries, necessitating substantial effort in typing and data entry.

It's essential to acknowledge that a significant proportion of civil engineering projects are situated in remote, challenging environments, often characterized by high altitudes or difficult-to-access locations. In these contexts, traditional access to computer systems becomes arduous. Moreover, users within the civil engineering domain may encounter difficulty in entering queries manually, particularly in remote settings where typing is cumbersome.

In step with the continuous evolution of computing devices, users are increasingly inclined toward a more seamless and efficient interaction with their technology. A diverse array of techniques and gadgets has been introduced to facilitate user- computer interactions, ranging from traditional mechanical devices like keyboards and mice to contemporary touch screens, motion capture systems involving gestures, and the integration of natural language input, including speech recognition. This drive for enhanced user experience and accessibility underscores the need for innovative solutions that can effectively bridge the gap between civil engineering professionals and complex standards and codes, particularly in challenging environments.

In the field of computing, the utilization of speech as an input has gained prominence. In this process, a computing device often transmits audio data to a remote computing system for speech processing. As users grow accustomed to harnessing voice commands to control their devices, conventional tasks can be reimagined to be more efficiently executed through voice-driven commands.

Electronic devices leverage a range of applications to respond to user speech input. For instance, when the speech input involves a command or a query, these devices employ various applications, often integrating advanced Al techniques described above. The result is a tailored response to the user, conveying the outcome of the executed operation. However, it's important to note that the accuracy, success rate, processing speed, and other aspects of these responses may vary depending on the specific application, introducing a degree of diversity and adaptability to the user experience.

As mobile devices continue to proliferate, users are increasingly dependent on these devices to access search engines and browse the web for their inquiries. However, despite the ubiquity of mobile device usage, many search engines and similar platforms have yet to offer interfaces and applications tailored for the extraction of essential data from civil engineering codes and the assessment of their interdependencies .

Accordingly, there is a need for a voice assisted device for object updating and query processing in the civil engineering domain. There is also a need for a method that provides an accurate and interactive response to a query received from a user.

SUMMARY OF THE INVENTION:

In one aspect, the present invention discloses a voice assisted device for object updating and query processing in the civil engineering domain for receiving an input query from a user, processing the input query and providing a machine generated response. The device of the present invention includes a processing unit equipped with a controller, a query module, processing engine, and a response module.

The query module is configured for receiving the input query in the form of one or more primary objects or secondary objects. The processing engine is configured for analyzing the input query, processing the received query, generating a plurality of embedded objects, computing the similarity among the embedded objects and stored objects, evaluating the output and generating the response. The response module is configured for providing the response in the form of one or more primary objects or secondary objects or tertiary objects.

The device of the present invention is operable in three modes namely: a first query mode, a second twin-tier mode, and a third multi-tier mode. The first query mode enables to receive the input query from the user in the form of primary objects and to generate the response in the form of primary or secondary objects. The second twin-tier mode enables to receive the input query from the user in the form of secondary objects and to generate the response in the form of secondary objects. The third multi-tire mode enables to receive the input query from the user in the form of primary or secondary objects and to generate the response in the form of tertiary objects.

The query module further includes a recorder and an input obtainer. The recorder is configured on the processing unit for registering the users for availing the facilities of device of the present invention and managing the access of the users. The input obtainer is configured on the processing unit for receiving the input query from user and for the query determination.

The processing unit is configured with processing engine that further includes modules namely an input processing module, an extraction module, a language processing module, a computation module, an output provider module, a learning and training module, a model updater module, and a knowledge base module.

The input processing module is configured on the processing unit for receiving the input, filtering the input data, and converting the input into the primary objects or secondary objects. The extraction module is configured on the processing unit for generating the embedded objects of input query by capturing their semantic and contextual meaning in a continuous vector space. The language processing module is configured on the processing unit for processing and optimizing the metadata for recognition the requirement of the user. The computation module is configured on the processing unit for computing the similarity between the embedded objects with the objects stored in the knowledge repository. The learning and training module is configured on the processing unit for training the processing engine for object identification, updating, and matching. The model updater module is configured on the processing unit for modifying the learning model based on evaluation of the recognition result provided by the output provider module.

The response module further includes a display module and a voice module. The display module displays the response in the form of primary objects. The voice module represents the response in the form of secondary objects or tertiary objects. In another aspect the present invention discloses a method performed by a voice assisted device for object updating and query processing receiving an input query from a user, processing the input query and generating a response, the method comprising the steps of: receiving an input query; obtaining features from a plurality of primary objects or secondary objects corresponding to the input query; obtaining metadata for the input query based on the obtained primary objects or secondary objects; processing the metadata for recognition and understanding the requirement of the user based on the metadata; fetching the matching data with respect to the recognized parameter of the requirement of the user; and outputting the response to the input query.

BRIEF DESCRIPTION OF DRAWINGS: The objectives and advantages of the present invention will become apparent from the following description read in accordance with the accompanying drawings wherein:

FIG.1 shows a block diagram illustrating a voice assisted device for object updating and query processing in accordance with the present invention;

FIG. 2 shows a schematic of components utilized for the voice assisted device for object updating and query processing of FIG. 1;

FIG. 3 shows a schematic of a processing unit of the voice assisted device for object updating and query processing of FIG. 1;

FIG. 4 shows various steps involved in operation of the voice assisted device for object updating and query processing of FIG. 1; and

FIG. 5 shows graphical representation illustrating an example in which a voice response is provided to a speech input according to the voice assisted device for object updating and query processing of FIG. 1.

DESCRIPTION OF THE INVENTION:

References in the specification to "one embodiment" or "an embodiment" means that a particular feature, structure, characteristic, or function described in connection with the embodiment is included in at least one embodiment of the invention. The appearances of the phrase “in one embodiment” in various places in the specification are not necessarily all referring to the same embodiment.

References in the specification to “preferred embodiment” means that a particular feature, structure, characteristic, or function described in detail thereby omitting known constructions and functions for clear description of the present invention.

The foregoing description of specific embodiments of the present invention has been presented for purposes of illustration and description. They are not intended to be exhaustive or to limit the present invention to the precise forms disclosed and obviously many modifications and variations are possible in light of the above teaching.

In general aspect, the present invention discloses a voice assisted device for object updating and query processing. The present invention provides real time and interactive electronic device for processing an input query and generating a response in the civil engineering domain.

Referring to FIG.l, a voice-assisted device 100 designed for object updating and query processing, denoted as "device 100," in accordance with the present invention. The device 100 receives an input query 105 from a user, processes the input query 105 and generates a response 110. Accordingly, the device 100 operates by receiving an input query 105 from a user, processing the query, and subsequently generating a tailored response, presented as 110.

The device 100 receives the request in various formats from the users. The device 100 includes a processing engine that includes a plurality of modules to process requests from the plurality of users for generating an output. The device 100 is the handheld device configured to communicate with the users via the input query 105 and the response 110. The device 100 advantageously includes the capabilities of machine learning and artificial intelligence, to proficiently execute tasks related to object updating and query processing. In accordance with the present invention, the integration of cutting-edge technology enhances the device's adaptability and efficacy in addressing user queries and facilitating seamless object updates. The device 100 follows a dynamic approach by incorporating each new user request as an addition to its dataset. The device 100 synchronizes its dataset with updated information at regular intervals, ensuring swift and precise query processing results. The continuous synchronization allows the device 100 to provide up-to-date and accurate responses to user queries.

In one embodiment, the device 100 takes diverse forms, serving as a tool for receiving input queries and delivering responses in multiple formats. The device 100 seamlessly adapts to user preferences, ensuring that responses are provided in formats such as text, voice, visual, audio-visual displays, etc. This versatility enhances the user experience, making information accessible in the most convenient and effective manner.

The device 100 of the present invention is a dedicated computing device with a storage capacity, one or more microprocessors, and one or more high speed network connections. In this one embodiment, device 100 assumes various forms, functioning as a versatile tool for receiving input queries and delivering responses in a range of formats. The device 100 intelligently adapts to user preferences, offering responses in text, voice, and visual displays. This adaptability optimizes the user experience, ensuring that information is accessible in the most convenient and effective manner. In context of another embodiment, the device 100 is any portable computing device that is carried by the user. Hereinafter, for convenience of explanation, a case where the device 100 is a dedicated handheld device is described as an example.

Referring to an embodiment of the present invention, the device 100 is operable in three modes namely a first query mode, a second twin-tier mode and a third multi-tier mode. In the first query mode, the input query 105 is receivable in the form of primary objects from the user and response 110 is generated in the form of primary or secondary objects. The primary object is a digital object, for example, a file in text format and the secondary object is a digital object, for example, a file in audio format.

In the second twin-tier mode, input query 105 is receivable in the form of secondary objects and response 110 is also in the form of secondary objects. In the third multi-tier mode, input query 105 is receivable in the form of primary or secondary objects and response 110 is in the form of tertiary objects. The tertiary object is a digital object, for example, a composite file featuring multi-media. The output is one or more digital objects that provides analysis of various parameters of the received input query 105.

In this specific embodiment of the present invention, the operation of device 100 is within the context of its application in the civil engineering industry. Nonetheless, it's essential to underscore that device 100 is not limited solely to this application. It exhibits versatility and can be effectively employed in other domains, including virtual education systems and project management systems, among others.

Now referring to FIG. 2, a schematic of the device 100 is discussed. The device 100 has a processing unit 205 and a controller 235. The processing unit 205 includes the controller 235 that is configured to execute one or more stored instructions for query processing. The controller 235 includes modules namely a query module 240, a processing engine 245, and a response module 250.

The device 100 includes an input/output (I/O) interface 220 for communication with other devices 110. The VO interface 220 is connected to one or more VO devices 225. The device 100 also includes one or more network interfaces 230 to enable communications between the device 100 and other networked devices. Such network interfaces 230 may include one or more network interface controllers (NICs) or other types of transceiver devices configured to send and receive communications over the communication network. The device 100 also includes one or more buses or other internal communications hardware or software that allow the transfer of data between the various modules and components of the device 100.

The device 100 includes a memory 210. The memory 210 provides storage of computer readable instructions, data structures, program modules, and other data for the operation of query processing of the device 100. The knowledge repository 215 is configured to store and retrieve the data of the device 100. The knowledge repository 215 is advantageously stored and auto updated with the data created within the device 100. Referring to the preferred embodiment, the knowledge repository 215 includes a dataset of specifications, codes and standards required in civil engineering domain. The knowledge repository 215 also stores dataset of frequently asked questions and their answers with respect to current specifications, codes and standards. This comprehensive repository 215 defines a valuable resource for users seeking quick and precise information related to these regulations and guidelines.

Referring to FIG. 3, a block diagram of the processing unit 205 of device 100 is described. The processing unit 205 includes the query module 240, the processing engine 245, and the response module 250. The query module 240 manages user interaction and input reception in various forms, including primary and secondary objects. The processing engine 245 examines the input query, conducts query processing, and produces the resulting data. The response module 250 engages in user interaction, delivering responses in various formats, including through a built-in speaker in the device 100.

The query module 240 handles the interaction and communication with the user while receiving the inputs. The query module 240 may receive the input query 105 in the form of primary objects or secondary objects. For example, the user provides the voice input through a microphone provided in the device 100. The processing engine 245 analyses the input query, processes the received query and generates the output data. The response module 250 handles the interaction and communication with the user while providing the output. The response module 250 may provide the response in the form of secondary objects 110. For example, the response through a speaker provided in the device 100. The query module 240 further includes a recorder 305 and an input obtainer 310. The recorder 305 registers the users for availing the facilities of device 100 and manages the access of the users. The recorder 305 is configured with the authentication and subscription functionality. The input obtainer 310 receives the input query 105 from user. The input obtainer 310 gathers data necessary for the query determination. The input obtainer 310 obtains the plurality of objects that are used for processing the input query.

The processing engine 245 is implemented with an input processing module 315, an extraction module 320, a language processing module 325, a computation module 330, an output provider module 335, a learning and training module 340, a model updater module 345 and a knowledge base module 350.

The input processing module 315 receives the input received from the input obtainer 310. The data received from the input obtainer 310 in the form of primary object or secondary object is processed, filtered, and then transmitted by the input processing module 315 to the extraction module 320. If the input query is received in the secondary object format, the input processing module 315 converts it into primary objects and then filters the converted primary object and then transmits it to the extraction module 320.

The extraction module 320 receives the filtered input data from the input processing module 315 and generates embedded objects. The embedded objects are fixed-length vector representations of input query that capture their semantic and contextual meaning in a continuous vector space. The extraction module 320 stores the embedded objects as a metadata. The embedded objects stored in metadata are computed as vectors in a high-dimensional vector space, where each dimension represents a feature or characteristic of the object.

The metadata additionally includes a plurality of features extracted from the input query 105 as well as the keyword extracted from the primary objects corresponding to the input query 105. For example, the metadata may include a variety of information that may be obtained from the text, such as the keyword extracted from the text corresponding to the input query 105, information about a sound characteristic of the input query 105, information about the user of the device 100 that receives the input query 105, etc., as information related to the input query 105 besides the text.

The language processing module 325 utilizes metadata from the extraction module to recognize and comprehend user requirements, optimizing metadata in vector space to distinguish semantically similar and dissimilar objects via a contrastive loss function. The language processing module 325 receives the metadata from the extraction module. The language processing module 325 processes the metadata for recognition and understanding the requirement of the user. The received metadata for each input query 105 is optimized to be similar in vector space for semantically similar objects and to be dissimilar for dissimilar objects, based on a contrastive loss function.

The computation module 330 determines object similarity by analyzing the semantic and contextual meaning of sentences, documents, or words within the objects and comparing them to those stored in the knowledge repository 215. In accordance with the present invention, the computation module 330 advantageously includes artificial intelligence functionality that fetches requested information with respect to the recognized requirements from the knowledge repository 215.

The computation module 330 computes the similarity between the embedded objects with the objects stored in the knowledge repository 215. The computation module 330 computes the similarity by calculating the similarity between two sentences, documents, or words within the objects based on their semantic and contextual meaning. It is to be noted, however, that sentence similarity among objects is computed using cosine similarity, and the output provider module 335 produces a response that includes the outcome of executing an operation based on the user's input query 105. In this one embodiment, the cosine similarity is used to calculate sentence similarity among objects. The output provider module 335 generates a response including a result of performing an operation according to the input query 105 of the user.

The learning and training module 340 includes machine learning models built on training datasets, serving to train the processing engine for object identification, updating, and matching, by utilizing acquired data to learn and establish references for accurately discerning user requirements. The learning and training module 340 includes one or more machine learning models created using the training datasets. The learning and training module 340 is implemented to train the processing engine for object identification, updating, and matching. The learning and training module 340 obtains data to be used for learning and applies the obtained data to a learning model thereby learning the reference for determining the exact requirement of user. The model updater module 345 adjusts the learning model by assessing the recognition results supplied by the output provider module 335 and shares this feedback with the learning and training module 340 to facilitate learning model modifications. The model updater module 345 modifies the learning model based on evaluation of the recognition result provided by the output provider module 335. For example, the model updater module 345 provides the learning and training module 340 with the recognition result provided by the output provider module 335 such that the learning and training module 340 modifies the learning model.

The knowledge base module 350 defines a comprehensive repository that encompasses a wide array of specifications, standards, and codes pertinent to the domain of civil engineering projects. It acts as a centralized and organized resource for accessing the essential guidelines, requirements, and regulations that govern civil engineering practices. Referring to the preferred embodiment of the present invention, the device 100 assists the engineering team in interpretation of the codes, standards and specifications required in civil engineering projects in prevailing country or area. The device 100 enables users to verify their specifications/code provisions before approval of the project and to deliver infrastructure projects efficiently.

The response module 250 engages in user interaction, delivering responses in various formats, including through a built-in speaker in the device 100. The response module 250 further includes a display module 355 and a voice module 360. The response module 250 is configured for providing the response 110 in the form of one or more primary objects or secondary objects or tertiary objects. The display module 355 displays the response 110 in the form of primary objects. The voice module 360 represents the response 110 in the form of secondary objects or tertiary objects.

Now referring to FIG. 4, a flowchart including various steps involved in the operation of the device 100 of the present invention is described. In the first step 405, the user accesses the user device 110.

In the next step 410, the access credentials are received from the user and the user authentication is performed. In the next step 415, the device 100 receives the input query 105. The input query 105 may include, for example, a speech command or text command for requesting an operation intended by the user.

In the step 420, the received data is processed by the input processing module 315 and the embedded objects are generated from the input query 105 by the extraction module 320. In the step 425, decoding of the extracted objects is performed by language processing module 325 and the decoded objects are transmitted to computing module 330 for further analysis.

In next step 430, the computing module 330 fetches requested information with respect to the recognized requirements by computing similarity between objects and evaluates the query. In the next step 435, the computing module 330 selects the optimal response for the input query 105. In next step 440, the output provider module 335 generates the response 110. In the next step 445, the response module presents the response 110 to the input query 105 according to the mode selected by user. In a step 450, the process is terminated for that instance. Referring to FIG. 5, a graphical representation illustrating an example in which a voice response is provided to a speech input according to the device 100 is discussed. The exemplary user interface 500 includes a first label 505, a second label 510, a third label 515 and a fourth label 520.

The first label 505 and the third label 515 display the first input query asked in voice format by the user. The second label 510 and the fourth label 520 show the response generated by the device 100 in voice format.

Now, referring to FIGS. 1 to 5, the operation of device 100 is described. The user accesses the device 100 by signing into the device 100. The credentials for accessing are received from the user device 100 and the user authentication is performed. The device 100 receives the access credentials from the user and the user authentication is performed. The device 100 receives the input query 105. The user provides the input query in the form of a speech command or a text command for requesting an operation intended by the user. The received data is processed by the input processing module 315 and the features are extracted from the input query 105 by the extraction module 320. Further, decoding of the extracted objects is performed by language processing module 325 and the decoded objects are transmitted to computing module 330 for further analysis.

The computing module 330 fetches requested information with respect to the recognized requirements by computing similarity between objects and evaluates the query. The computing module 330 selects the optimal response for the input query 105. The output provider module 335 generates the response 110. The response module presents the response 110 to the input query 105 according to the mode selected by user.

A person skilled in the art will appreciate that the operation of device 100 commences with user access, requiring authentication via the submission of access credentials from the user's device. Subsequently, the input query 105 is received with users opting for either speech or text commands to articulate their intentions.

The input query advantageously undergoes a series of processing steps: initial data processing through the input processing module 315, feature extraction performed by the extraction module 320, followed by object decoding facilitated by the language processing module 325. These decoded objects are then relayed to the computing module 330 for in-depth analysis. Within this analytical phase, the computing module 330 determines the most fitting response by measuring object similarity and evaluating the query. The output provider module 335 generates the response 110, which is presented to the user in their preferred mode of interaction.

It is to be noted that in the first embodiment of the present invention, the features of the device 100 are accessible to the users via web interfaces. In this first embodiment, the user devices are configured with a client component of a system for object updating and query processing for accessing the web interface of the present invention. The web interfaces on the user devices are controlled by a server of the system for object updating and query processing. The server and the user devices are connected to each other via a communication network. The users using their personal computing devices such as computer, laptop, and smart phone accesses the services and functions provided by the device 100 via the web interface supported by the server.

The client component of the above-mentioned system can be seamlessly configured as an application on any computer or tablet, offering users the flexibility to access its valuable features and functionalities on their preferred digital devices. Its adaptability across various computing platforms ensures a user-friendly experience, making it accessible wherever professionals or individuals require quick and efficient access to pertinent information in the field of civil engineering or other applications.

Now, a method performed by a voice assisted device and a system for object updating and query processing 100, receiving an input query 105 from a user, processing the input query 105 and generating a response 110 is discussed. The said method includes the steps of: a. receiving an input query; b. obtaining features from a plurality of primary objects or secondary objects corresponding to the input query 105; c. obtaining metadata for the input query 105 based on the obtained primary objects or secondary objects; d. processing the metadata for recognition and understanding the requirement of the user based on the metadata; e. fetching the matching data with respect to the recognized parameter of the requirement of the user; and f. outputting the response 110 to the input query 105. Accordingly, the method performed by the device 100 for object updating and query processing 100 is a comprehensive process that streamlines user interaction and delivers efficient responses. This method involves the following steps such as receiving User Input. The process commences by receiving an input query 105 from the user. For example, a user may input a query such as, "What are the safety standards for constructing bridges?".

In the next step of feature extraction, features are obtained from a variety of primary and secondary objects that correspond to the input query. These objects could include text, documents, or data. For instance, if the user's query pertains to bridge safety, the system collects data from relevant documents and codes on bridge construction safety.

Then metadata is obtained for the input query based on the information gathered from the primary and secondary objects. This metadata captures essential details about the user's query. In an example, it might include keywords like "bridge safety standards" and "construction regulations."

In the next step of recognition and understanding, the metadata is processed to recognize and understand the user's requirements. Using machine learning and natural language processing, the device interprets the user's intent. In the case, it comprehends that the user seeks information about safety standards in bridge construction.

Further data retrieval is done based on the recognized requirements. The device advantageously fetches matching data from its knowledge repository. It retrieves relevant information from its extensive database, such as the latest bridge safety standards.

In a step of response generation, the device 100 outputs the response 110 to the user's input query. It may provide the user with a detailed answer, summarizing the safety standards for bridge construction, and ensuring a swift and accurate response to the user's query.

In this way, the method ensures that the user receives precisely the information they need, saving time and effort while enhancing their interaction with the device.

The device of the present invention advantageously provides an adaptive and optimized environment for object updating and query processing. The present invention provides real time and interactive electronic device for query processing. The present invention advantageously provides more accurate, cost effective and efficient device for object updating and query processing based on machine learning and Al functionality.

The device 100 holds a notable advantage by promoting environmental friendliness. It does so by offering Indian code authorities a means to dispense with traditional hard or soft copies of codes and instead make them readily accessible through the device. In doing so, it contributes to reducing paper usage and conserving resources. This innovative approach serves as a valuable solution for engineers and professionals in the civil engineering industry. By granting swift and efficient access to relevant codes and regulations, the invention optimizes the user experience, saving significant time and effort that would otherwise be expended on manual searches through extensive volumes of information.

The invention presents a host of valuable advantages. Firstly, it facilitates seamless interaction by allowing users to communicate with the device via either speech or text commands, enhancing accessibility and user-friendliness. This adaptability extends to the device's ability to provide customized responses due to the intelligent analysis of user queries using machine learning models, ensuring that the information delivered precisely aligns with individual user requirements.

Furthermore, the device's has ability of providing real-time updates, achieved through the model updater module, which continually refines the learning model based on recognition results. As a result, the device evolves and improves its accuracy over time, ensuring it remains relevant and effective. Its versatile nature is a standout feature, initially designed for civil engineering applications but with potential utility across diverse domains, such as virtual education and project management.

Moreover, the device 100 boasts a centralized knowledge repository, housing a comprehensive collection of specifications, standards, and codes in the civil engineering field. This repository streamlines the access to essential information, making it an indispensable resource for professionals in the industry. Fast and accurate information retrieval is another significant advantage, achieved by computing object similarity and evaluating user queries, streamlining data retrieval processes. Users further benefit from the flexibility offered in response formats, including text, voice, and visual displays, allowing for a tailored and convenient user experience.

The embodiments were chosen and described in order to best explain the principles of the present invention and its practical application, to thereby enable others, skilled in the art to best utilize the present invention and various embodiments with various modifications as are suited to the particular use contemplated.

It is understood that various omission and substitutions of equivalents are contemplated as circumstance may suggest or render expedient, but such are intended to cover the application or implementation without departing from the scope of the present invention.