A system and method for safety management for construction sites

Abstract

The invention provides a safety management system for construction sites. The system comprises an inputting module adapted to receive an input from a user; a processing module adapted to process the input from the user, wherein the processing module comprises an extracting module adapted to extract information from the input, and a validating module adapted to validate the information extracted to form a validated information; and an outputting module adapted to output the validated information to a receiving terminal.

Description

1 A System And Method For Safety Management For Construction Sites Field of the Invention The invention relates to the field of construction safety management. Particularly but not exclusively, the invention relates to a mobile-based, artificial intelligence-enabled system and 5 method for managing workers’ safety at construction sites. Background of the Invention Construction sites are inherently high-risk environments requiring stringent safety monitoring. Traditional safety management often relies on top-down supervision, paper-based 10 records, or simple digital forms. These approaches are known to suffer from drawbacks and limitations. For example, frontline workers frequently show low engagement, as they often view safety reporting as "extra work" or a purely punitive measure, leading to underreporting of near- misses. Safety officers are often required to manually verify records and re-enter data into industry-wide systems, for example, the Frontline Personnel Safety Performance Recording 15 Scheme (FPSPRS) of the Construction Industry Council (CIC), which is tedious, error-prone and inefficient. Particularly, manual entry of worker credentials and certificates, such as Construction Industry Safety Training Certificate (so-called the Green Cards) or Construction Workers Registration Certificates (CWRC) is especially time-consuming and susceptible to inaccuracies. Moreover, existing workforce management tools generally lack integration across various 20 functionalities, such as site access control, safety performance tracking, other operational management. There is therefore a clear need for a system that improves safety compliance from a passive obligation into active participation, while automating data synchronization and reducing administrative burden. 25 Objects of the Invention An object of the present invention is to provide a system and a method for safety management for construction sites. Another object of the present invention is to mitigate or obviate to some degree one or more problems associated with known safety management systems for construction sites, or at 30 least to provide a useful alternative. HK30135235 A 2 The above objects are met by the combination of features of the main claims; the sub- claims disclose further advantageous embodiments of the invention. One skilled in the art will derive from the following description other objects of the invention. Therefore, the foregoing statements of object are not exhaustive and serve merely to illustrate some of the many objects of the present invention. 5 Summary of the Invention In a first main aspect, the invention provides a safety management system for construction sites. The system comprises an inputting module adapted to receive an input from a user; a processing module adapted to process the input from the user, wherein the processing module 10 comprises an extracting module adapted to extract information from the input, and a validating module adapted to validate the information extracted to form a validated information; an outputting module adapted to output the validated information to a receiving terminal. In a second main aspect, the invention provides a method of safety management for construction sites. The method comprises the steps of receiving an input from a user via an input 15 module; wherein the input comprises a text, an image, a video, a code, and / or an audio; processing the input via a processing module, wherein the step of processing comprises the steps of extracting, via an extracting module, information from the input; and validating, via a validating module, the information extracted to form a validated information; and analyzing, via an analyzing module, the validated information to determine a level of risk associated with the 20 validated information. The summary of the invention does not necessarily disclose all the features essential for defining the invention; the invention may reside in a sub-combination of the disclosed features. Brief Description of the Drawings 25 The foregoing and further features of the present invention will be apparent from the following description of preferred embodiments which are provided by way of example only in connection with the accompanying figures, of which: Fig. 1 is a block schematic diagram showing a safety management system for construction sites according to an embodiment of the present invention. 30 HK30135235 A 3 Description of Preferred Embodiments The following description is of preferred embodiments by way of example only and without limitation to the combination of features necessary for carrying the invention into effect. Reference in this specification to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is 5 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, nor are separate or alternative embodiments mutually exclusive of other embodiments. Moreover, various features are described which may be exhibited by some embodiments and not by others. Similarly, various requirements are described which may be 10 requirements for some embodiments but not other embodiments. It should be understood that the elements shown in the figure, may be implemented in various forms of hardware, software or combinations thereof. Preferably, these elements are implemented in a combination of hardware and software on one or more appropriately programmed general-purpose devices, which may include a processor, memory and input / output 15 interfaces. The present description illustrates the principles of the present invention. It will thus be appreciated that those skilled in the art will be able to devise various arrangements that, although not explicitly described or shown herein, embody the principles of the invention and are included within its spirit and scope. 20 Moreover, all statements herein reciting principles, aspects, and embodiments of the invention, as well as specific examples thereof, are intended to encompass both structural and functional equivalents thereof. Additionally, it is intended that such equivalents include both currently known equivalents as well as equivalents developed in the future, i.e., any elements developed that perform the same function, regardless of structure. 25 Thus, for example, it will be appreciated by those skilled in the art that the block diagrams presented herein represent conceptual views of systems embodying the principles of the invention. The functions of the various elements described or shown in the figures may be provided through the use of dedicated hardware as well as hardware capable of executing software in 30 association with appropriate software. When provided by a processor, the functions may be HK30135235 A 4 provided by a single dedicated processor, by a single shared processor, or by a plurality of individual processors, some of which may be shared. Moreover, explicit use of the term "processor", “processing device” or "processing module" should not be construed to refer exclusively to hardware capable of executing software, and may implicitly include, without limitation, digital signal processor ("DSP") hardware, read-only memory ("ROM") for storing 5 software, random access memory ("RAM"), and non-volatile storage. In the claims hereof, any element expressed as a means for performing a specified function is intended to encompass any way of performing that function including, for example, a) a combination of circuit elements that performs that function or b) software in any form, including, therefore, firmware, microcode or the like, combined with appropriate circuitry for 10 executing that software to perform the function. The invention as defined by such claims resides in the fact that the functionalities provided by the various recited means are combined and brought together in the manner which the claims call for. It is thus regarded that any means that can provide those functionalities are equivalent to those shown herein. The present invention relates to a system and a method for safety management within 15 defined environments, such as but are not limited to construction sites. Particularly, the invention integrates multi-modal artificial intelligence with internet-of-things (IoT) technologies to significantly improve the accuracy, timeliness and operational efficiency of on-site safety management. By automating capture, extraction, validation, and synchronization of safety data, the system minimizes manual data entry and reduces administrative errors and delays. For 20 example, by automating data synchronization including submission of validated records to external registries and regulatory schemes such as the Hong Kong Frontline Personnel Safety Performance Recording Scheme (FPSPRS) by CIC, and / or to other relevant authorities. The invention further provides a gamified incentive mechanism that rewards proactive safety behaviors, for example, timely reporting near-misses or completion of safety training. 25 Redeemable Reward Points can be awarded as incentives to encourage a sustained culture of safety rather than mere monitoring. Furthermore, the system enhances site security and risk mitigation through intelligent access control that automatically restrict, grant, or conditional permit workers access based on real-time safety performance and training records. The invention therefore provides a closed-loop "data capture-validation-intervention-incentive" workflow that 30 HK30135235 A 5 ensures higher data accuracy, streamlined regulatory compliance, improved on-site safety, and enhanced worker engagement, resulting in a safer and more efficient working environment. Without being limited by the described embodiments and the drawings, a skilled person in the art will appreciate that the present invention should not be restricted to applications in construction sites or for construction sites of any specific purposes. Instead, any venues, locations 5 or facilities, either indoor or outdoor, for use by any industries or applications, shall also be encompassed, as long as the applications do not depart from the inventive concept of the present invention. Referring to Fig. 1, shown is a schematic diagram illustrating system 10 for construction site safety management in accordance with an embodiment of the present invention. The system 10 10 may include a communication equipment such as smart phones, tablets or handheld personal computer equipped with specific application (APP) or software for communication with various modules of system 10, and / or be adapted to receive messages via remote communication network such as via short message service (SMS) or other communication application (APP) or software. The system 10 preferably comprises an inputting module 20 configured to receive an input from a 15 user. In one embodiment, the inputting module 20 comprises one or more of a text inputting device 22, an image capturing device 24, a video capturing device 26, a code scanning device 30, and a voice capturing device 28, for receiving input in the form of texts, images, videos, machine readable or scannable codes, and / or audios, respectively. The inputting module 20 may further comprise other inputting means 32 for receiving alternative types of input signals from the user, 20 such as biometric data, for example. The system 10 further comprises a processing module 40 configured to process the received user input. In one embodiment, the processing module 40 comprises an extracting module 42 adapted to extract information from the user input, and a validating module 44 adapted to validate the extracted information. In one embodiment, the extracting module 42 25 comprises one or more of an optical character recognition (OCR) module 43, a natural language processing (NLP) module 45 and / or other processing components executing an artificial intelligence (AI) machine learning algorithm. Preferably, the optical character recognition (OCR) module 43 and / or a natural language processing (NLP) module 45 are equipped with artificial intelligence capacities. For example, the OCR module 43 may provide image-to-text conversion 30 and structured data extraction from the user input, processing visual artifacts encountered during HK30135235 A 6 onboarding registrations and site operations. OCR module 43 converts inputs such as photographs, scanned documents, identity cards, training certificates, permits, safety signage, equipment labels, and other image-based records into machine-readable text and enters the extracted information into structured fields. Extracted data such as worker profiles, certification records, incident reports, and regulatory submissions can be automatically populated into 5 required formats, along with verification of credentials for synchronization with external systems, such as the FPSPRS of the CIC, for example. The NLP module 45 provides language understanding and structured information extraction capabilities across the system 10. The NLP module 45 converts unstructured human language, such as worker reports, chat messages, safety descriptions, and regulatory guidance into machine‑readable, structured representations usable by 10 downstream components. For example, the NLP module 45 may support tasks such as intent detection, entity extraction, hazard classification, sentiment or urgency estimation, and normalization of user inputs to the system’s internal data schema. The modular processing capability therefore provides not only accurate content extraction but also contextual analysis of that content. 15 In one embodiment, the validating module 44 may comprise one or more components such as a large language model (LLM) module 47, a retrieval-augmented generation (RAG) module 49, and / or other components executing an artificial intelligence (AI) machine learning algorithm. Preferably, the large language model (LLM) module 47 and the retrieval-augmented generation (RAG) module 49 are equipped with application-specific artificial intelligence 20 capacities. For example, the LLM module 47 functions as the system’s reasoning and synthesis engine that evaluates structured data produced by the extracting module 42, generates validation verdicts and human-readable rationales, and formulates targeted clarification prompts or recommended suitable actions, while relying on retrieved evidence to avoid unsupported assertions. The RAG module 49 provides the authoritative, relevant evidence such as documents 25 or passages that grounds the LLM validation decisions. The RAG module further locates, ranks, and supplies concise, high‑trust content that the LLM module 47 can rely upon when validating the information or data. Merits of the present invention can be further explained and elaborated based on the following example. In one specific embodiment, the system 10 can be provided in the form of a 30 mobile application (APP) installable on a worker’s mobile device or other computing devices. HK30135235 A 7 During an on-boarding registration or daily check-in, the worker utilizes the image capturing functionality 24, or other input interfaces of the inputting module 20 to capture images of the relevant documents, such as but are not limited to, identity cards, work certificate, Construction Industry Safety Training Certificate (the “Green Cards”) or Construction Workers Registration Certificates (CWRCs). The AI-enhanced OCR Module 43 and / or NLP module 45 perform 5 contextual recognition of the information and convert these images and inputs into normalized, structured data. The data will then be validated by the LLM module 47 in conjunction with the evidence retrieved by the RAG module 49 to confirm document authenticity, certificate validity and contextual consistency. Validated records will be automatically formatted and populated in the required fields in specific formats. Through the outputting module 80, the system 10 securely 10 submits the formatted data to a receiving terminal 100, which can be one or more of a local or remote databases, site management systems, regulatory servers, or systems of authoritative bodies such as the FPSPRS of the CIC, for example. By automating the capture, extraction, evidence‑grounded validation, and submission of worker credentials and related records, the AI‑powered extracting module 42 and validating module 44 significantly reduces manual data 15 entry, eliminates duplicate transcription, improves communications between workers and site managers, accelerates synchronization with regulatory authorities, and improves overall data accuracy and integrity. In one further embodiment, the system 10 may optionally comprise an access control module 50 which controls access of the user to one or more designated zones or regions of the 20 construction site based on successful validation and submission of the required documents, such as when a worker’s Green Card and / or CWRC have been authenticated. The system may further comprise an AI-driven chatbot module 48 to provide workers with convenient, on-demand access to safety and environmental guidance. Particularly, the chatbot module 28 supports natural- language queries about safety procedures and regulatory requirements, identifies potential 25 hazards via consulting an integrated knowledge base, recommends mitigation steps, and / or delivers real-time project updates alerts which may include, for example, changes to work schedules, site including severe weather conditions and when Safety Performance point reach 15 and pollution conditions, and safety briefings, etc. The access‑control and chatbot capabilities enable conditional, evidence‑backed site entry and provide timely, actionable support to frontline 30 personnel. HK30135235 A 8 Preferably, the system 10 further comprises an analyzing module 60 adapted to analyze the validated information and determine a level of risk associated with the validated information. An assigning module 70 is then employed to assign a score according to the determined level of risk associated with the validated information, with the outputting module 80 further transmitting the score to the receiving terminal 100. The scores can be assigned in respect of the user, to any 5 person related to or identified in the validated information, or both, according to the determined level of risk. For example, Worker A may observe or identify an unsafe behavior or action by Worker B, for example, worker B not wearing a safety helmet in the construction site. Worker A can use the system 10 mobile APP and the image capturing device 24 to take a photo of the unsafe act 10 and submits a report via the system 10. Worker A may optionally supplement the report with text description and / or audio to identify the unsafe action. The analyzing module 60, incorporated with AI-equipped functionality, processes and analyses the image, text, and / or audio to identify the unsafe action and determine its risk level. The assigning module 70 then assigns a score, which can be a positive score or negative score, based on the determined risk. If Worker B’s 15 behavior is judged risky, Worker B might receive point for example, a demerit point (for example, -10). Conversely, Worker A, for timely reporting the unsafe behavior, might earn positive merit point (for example, +5). The nature and allocation of points can be predetermined by the system administrators, and a person skilled in the art would appreciate that variations in terminology, scales, allocation methods and calculation techniques, etc. are possible. The 20 outputting module 80 transmits assigned scores to the receiving terminal 100 in real time. If a worker’ accumulative scoring falls below a configurable negative threshold, indicating repeated unsafe behavior, the access module 50 is triggered to block that worker’s entry to specified zones or to the construction site. Conversely, if a worker completes approved safety-improvement courses and registers the corresponding certificates with system 10, the 25 worker is awarded with points. Once reward points reach a configurable positive threshold, the access module 50 restores the worker’s access to previously restricted areas. The system 10 implements strict data-privacy, authentication, and anti-abuse safeguards to protect users and ensure report integrity. All submitted reports, including images, audio, and text and derived scores are transmitted and stored using end-to-end encryption and retained only 30 for configurable retention periods to meet legal and organizational requirements. To reduce false HK30135235 A 9 or malicious reports, the analyzing module 60 applies automated verification, such as cross- referencing report metadata on location, time, device ID, visual and / or audio content validation, and corroboration with other sensors or site data to flag low-confidence or conflicting submissions for manual review by authorized safety personnel. The system 10 also supports reputation weighting, for example, adjusting score impact based on reporter reliability, rate 5 limiting and anomaly detection to limit spam or targeted abuse. In one embodiment, a worker’s scores are credited in real time to the worker’s electronic wallet or profile within the system 10. The system 10 may further comprise a redemption module 90 that converts all or part of a user’s accumulated score into rewards. Rewards may take different forms, for example, a redeemable code, such as a QR code that can be scanned at 10 vending machines 110 or participating merchants such as retail shops, canteens, cafes or restaurants to obtain consumer items, e.g. a drink, food, vouchers, gift items, or services. Reward may also be delivered as payroll-linked benefits, credits toward company programs or other incentives defined by the operator. Users can manage earned scores and rewards securely from the system’s web or mobile interface, such as to view a detailed redemption history or transaction 15 balance. In a further aspect of the present invention, it provides a method of safety management for construction sites implemented the system 10 as described above. The method comprises receiving an input from a user via the input module 20, where the input may comprise a text, an image, a video, a code, and / or an audio; processing the input via a processing module 40, wherein 20 the step of processing comprises the steps of extracting, using the extracting module 42, information from the input; and validating, using the validating module 44, the information extracted to form a validated information. The method further comprises analysing, via the analysing module 60, the validated information to determine a level of risk associated with the validated information, and assigning, via the assigning module 60, a score according to the 25 determined level of risk; and transmitting the assigned score to the receiving terminal 100 via an outputting module 80. Optionally, the method further comprises a step of converting all or part of the accumulated score into a reward via a redemption module 90. The invention is advantageous in that it integrates multi-modal artificial intelligence (AI) with internet-of-things (IoT) technologies to significantly improve accuracy, timeliness and 30 operational efficiency of on-site safety management in construction and other high-risk HK30135235 A 10 environments. By combining multi-modal data capture, automated information extraction, validation and regulatory synchronization, as well as real time reporting of unsafe behaviors or situations, the invention minimizes manual data entry, reduces administrative overhead, and automates safety synchronization with authoritative registries. The invention further provides a gamified incentive mechanism that rewards proactive safety behaviors, for example, issues 5 reward points in real time for reporting near-misses. The reward points are redeemable through the system, thereby encouraging a sustained culture of safety rather than mere monitoring. Furthermore, the system enhances site security and mitigates risks through intelligent access control. Workers can be automatically blocked or unblocked from site zones based on their real- time safety performance and training records. The invention therefore provides a closed-loop 10 "data capture-validation-intervention-incentive" workflow that ensures higher data accuracy, supports regulatory compliance, and fosters a safer working environment. The present description illustrates the principles of the present invention. It will thus be appreciated that those skilled in the art will be able to devise various arrangements that, although not explicitly described or shown herein, embody the principles of the invention and are included 15 within its spirit and scope. Moreover, all statements herein reciting principles, aspects, and embodiments of the invention, as well as specific examples thereof, are intended to encompass both structural and functional equivalents thereof. Additionally, it is intended that such equivalents include both currently known equivalents as well as equivalents developed in the future, i.e., any elements 20 developed that perform the same function, regardless of structure. While the invention has been illustrated and described in detail in the drawings and foregoing description, the same is to be considered as illustrative and not restrictive in character, it being understood that only exemplary embodiments have been shown and described and do not limit the scope of the invention in any manner. It can be appreciated that any of the features 25 described herein may be used with any embodiment. The illustrative embodiments are not exclusive of each other or of other embodiments not recited herein. Accordingly, the invention also provides embodiments that comprise combinations of one or more of the illustrative embodiments described above. Modifications and variations of the invention as herein set forth can be made without departing from the spirit and scope thereof, and, therefore, only such 30 limitations should be imposed as are indicated by the appended claims. HK30135235 A 11 In the claims which follow and in the preceding description of the invention, except where the context requires otherwise due to express language or necessary implication, the word “comprise” or variations such as “comprises” or “comprising” is used in an inclusive sense, i.e. to specify the presence of the stated features but not to preclude the presence or addition of further features in various embodiments of the invention. 5 HK30135235 A 1 Claims 1. A safety management system for construction sites, comprising: an inputting module adapted to receive an input from a user; a processing module adapted to process the input from the user, wherein the processing 5 module comprises an extracting module adapted to extract information from the input, and a validating module adapted to validate the information extracted to form a validated information; and an outputting module adapted to output the validated information to a receiving terminal. 10 2. The safety management system according to claim 1, further comprising an analyzing module adapted to analyze the validated information and to determine a level of risk associated with the validated information. 3. The safety management system according to claim 2, further comprising an assigning15 module adapted to assign a score according to the determined level of risk associated with the validated information; wherein the outputting module is adapted to transmit the score assigned to the receiving terminal. 4. The safety management system according to claim 3, wherein the input comprises a text,20 an image, a video, a code, and / or an audio. 5. The safety management system according to claim 3, wherein the inputting module comprises one or more of a text inputting device, an image capturing device, a video capturing device, a code scanner and / or a voice capturing device. 25 6. The safety management system according to claim 3, wherein the extracting module comprises one or more of an optical character recognition (OCR) module and / or a natural language processing (NLP) module. 30 HK30135235 A 2 7. The safety management system according to claim 3, wherein the validating module comprises one or more of a large language model (LLM) module, a retrieval-augmented generation (RAG) module. 8. The safety management system according to claim 3, wherein the score is assigned in 5 respect of the user and / or a person related to the validated information, according to the determined level of risk associated with the validated information. 9. The safety management system according to claim 8, further comprising a redemption module adapted to convert the score or part of an accumulated scores of the user and / or the 10 person into a reward. 10. The safety management system according to claim 9, wherein the reward comprising a code detectable at machines or merchants for redemption of a consumer goods or services. 15 11. The safety management system according to claim 8, further comprising an access control module to control access of the user and / or the person to one or more selected zones or regions of the construction sites. 12. The safety management system according to claim 1, wherein the outputting module is20 adapted to transmit the validated information in a specific format to the receiving terminal. 13. The safety management system according to claim 1, wherein the receiving terminal comprises one or more of a local database, a remote database, a local device, and / or a remote server. 25 14. A method of safety management for construction sites, comprising the steps of: receiving an input from a user via an input module; wherein the input comprises a text, an image, a video, a code, and / or an audio; processing the input via a processing module, wherein the step of processing comprises 30 the steps of: HK30135235 A 3 extracting, via an extracting module, information from the input; and validating, via a validating module, the information extracted to form a validated information; and analysing, via an analysing module, the validated information to determine a level of risk associated with the validated information. 5 15. The method according to claim 14, further comprising a step of assigning, via an assigning module, a score according to the determined level of risk associated with the validated information; and transmitting, via an outputting module, the score assigned to a receiving terminal. 10 16. The method according to claim 15, further comprising a step of converting, via a redemption module, the score or part of an accumulated scores into a reward. 15 20 25 30 HK30135235 A Inputting Module 20 RFID Module 24 Restaurants, Vending Machines or Other Merchants 110 Processing Module 40 Texting device 22 Image Capturing Device 24 Video Capturing Device 26 Video Capturing Device 28 FIG. 1 Scanning Device 30 Other Inputting Means 32 Extracting Module 42 Validating Module 44 OCR 43 NLP 45 LLM 47 RAG 49 Analysing Module 60 Assigning Module 70 Outputting Module 80 Cloud Database Mobile Devices Database Computing Devices Redemption Module 90 10 Access Control Module 50 Receiving Terminal 100 Chatbot Module 48 1HK30135235 A

Images