TECH OFFERS

Traditionally, Design for Manufacturing (DFM) feedback is provided late in the product development cycle, often resulting in costly redesigns, production delays, and missed opportunities for optimization. This invention introduces an AI-powered companion that delivers real-time DFM feedback early in the design and demand fulfillment process. By automatically analyzing CAD models and design parameters, it identifies features that may pose manufacturing challenges and offers actionable suggestions to improve feasibility and reduce costs. The AI companion empowers users to make informed design decisions without waiting for manual reviews or relying solely on in-house manufacturing experts. In additional, traditional quoting methods are labor intensive and inaccurate, requiring engineers to analyze technical drawings, assess material and process requirements, and estimate costs based on historical data or supplier inputs. This delays customer responses and increases operational overhead. The proposed technology leverages AI to automatically interpret design files (e.g., CAD or technical drawings), extract key manufacturing features, and generate instant, data driven price estimates.  The technology owner is looking for partners to codevelop this AI module and customise it to specific industry, application and business process. Target users include precision engineering firms, job shops and manufacturers. The solution can enable businesses to respond faster to customer inquiries, reduce quoting costs, reduce design for manufacturing leadtime and win more business.   AI Companion for DfM Validation: User will load in CAD or drawing. Solution will be based on unique requirements and manufacturing capabilities of customers. It calls out design constrains and offers actionable suggestions to improve feasibility and reduce costs. Automated Instant Pricing for Manufacturing: User will load in CAD or drawing and manufacturing process. Solution will project an estimated quotation with use cases at >80% accuracy. The AI engine calculates pricing by factoring in labor, material costs, machine time, overhead, and historical quotation data to ensure accurate and consistent estimates. Quotation can include estimated raw material cost, cost price and selling price.  Flexible and Integrated Deployment: Used as a plug-in API as part fo a larger internal tool, or a standalone solution with a simple UI developed for customer to upload and view design feedback. A web-based user interface and API allow users to upload design files, receive instant quotes, and seamlessly integrate the system into existing ERP, MRP, or e-commerce platforms. AI Powered: Technology consists of a combination of software algorithms, machine learning models, and integration tools. Supporting this is a robust database that stores material pricing, manufacturing parameters, and user interaction data, enabling continuous model learning and optimization over time. Scalable Implementation with Additional Features: DfM Validation and Instant Pricing solutions can be part of a plug-and-play AI module that cover 50 features across operation efficiency such as production job tracking and inventory management. Target users include precision engineering firms, job shops, manufacturers, online marketplace and any product owners. This includes users employing different manufacturing processes such as CNC machining, 3D printing, and sheet metal fabrication.  Solution is applicable across different industries including consumer, defense, aerospace and semiconductors. It is especially valuable in industries where rapid prototyping, custom part production, or low-volume manufacturing are common to achieve goals below: Real-time DFM feedback early in the design and demand fulfillment process to reduce number of design iterations required, shortens development timelines. Real-time manufacturability and cost feedback during the design phase. Backend engine for procurement automation, enabling sourcing teams to quickly benchmark and compare supplier quotes. The Computer-Aided Manufacturing (CAM) Market is expected to increase from $3.39 billion in 2024 to $5.69 billion by 2030, at a CAGR of 9.0%. This reflects big growth in the manufacturing sector. With so much to manufacture and design, it becomes even more pressing to create more efficient workflows, develop shorter design iteration cycles and get feedback in real time. In addition, the global manufacturing industry is a substantial and growing sector, valued at approximately USD 14.16 trillion in 2024 and projected to reach USD 20.76 trillion by 2031, with a compound annual growth rate (CAGR) of 4.9%. Within this vast landscape, the High-Mix, Low-Volume (HMLV) segment is emerging as a critical area of growth, driven by increasing demand for customisation, rapid prototyping, and shorter product life cycles across industries such as aerospace, medical devices, and electronics. Despite its strategic importance, the HMLV segment remains largely underserved, particularly in the area of pricing automation. Most job shops and small-batch manufacturers still rely on time-consuming, manual quoting processes for each unique part or drawing. This creates a significant bottleneck, limiting responsiveness, efficiency, and scalability. This solution stands out from similar technologies in the market because of the solutions providers strong record in servicing diverse clients. This tool has been verified and trained on real world design files from a variety of clients from different industries and has become a key portion of our optimised workflow. It is customised to customer's problem statement, supporting unique and niche designs for specific markets. The solution offers a powerful combination of CAD feature recognition, AI-driven cost estimation, and seamless ERP/API integration to automate pricing at speed and scale. Unlike existing tools like basic quoting templates, this system intelligently learns from historical data and adapts to new designs, delivering both accuracy and agility. Reduction in number of design iterations: Shortens development timelines, and supports the creation of more practical, production-ready designs. Significant reduction in turnaround time for quotation: Replaces traditional quoting methods are labor intensive and inaccurate, requiring engineers to analyze technical drawings, assess material and process requirements. Shorten quotation timeline from a few days to few seconds.  Auotmates workflow that typically requires profession: Provides analysis of CAD/drawings which is particularly beneficial for product design teams, manufacturing engineers, and procurement professionals who play a role in assessing manufacturability and sourcing.  For small and mid-sized manufacturers, this can lead to a higher quote win rate and better customer service. For larger enterprises, it supports scalability across multiple product lines or factories. manufacturing, operations, efficiency, ai, automation, process Infocomm, Artificial Intelligence, Manufacturing, Assembly, Automation & Robotics

Platinum group metals (PGMs) are critical raw materials essential in diverse industries, including automotive catalytic converters, jewelry, glassware, petrochemical refining, electronics, and healthcare sectors like pharmaceuticals and dental implants. Primarily sourced through the mining of PGM ores, they constitute about 70% of the global PGM supply, with South Africa and Russia accounting for 85% of this production. This concentration in supply can lead to price gouging and market monopoly. Recycling PGMs from waste not only mitigates the supply shortfall but also reduces environmental impacts compared to mining. However, conventional recycling methods are energy-intensive, requiring temperatures around 1500°C, and involve costly downstream processing to treat waste. Furthermore, the high processing temperatures result in high-value raw materials being burnt and releasing harmful toxins. The technology owner has developed a novel biorecovery method that incorporates and modifies a series of biochemical and biological processes into a streamlined 3-stage process as opposed to the multi-tiered stages of current conventional methods used in industry. It offers the following advantages over the competition: Energy Efficiency: consumes 6x less energy than traditional methods Cost Effective: 3x cheaper in operation cost High Yield: capable of recovering multiple PGM simultaneously with high yield even from low-grade waste Sustainability: support company decarbonization goals by offering a truly green and sustainable recycling manner for spent catalyst The core process and specifications of the technology are summarised as follows: Statistically-Optimised Ultrasonication: as a key pretreatment step, this sonication method effectively removes all undesirable metals from waste, isolating PGM-rich materials, called the PGM-preconcentrated stream, enhancing the efficiency of subsequent steps. Bioextraction Technique: secondly, utilise a novel and unique bioextraction technique to extract PGMs from waste with high efficiency (i.e., 99% recycling rate per cycle for rhodium (Rh), 92-95% per cycle recycling rate for platinum (Pt) and palladium (Pd)). It can be employed at a commercial scale without compromising yield. Bioreduction, Bioaccumulation, and Bioprecipitation: a combination of these improved biological processes are used in the third step to produce PGM into powder form which further undergoes separation and purification to produce high-purity PGM products. This technology is ideal for industries that are interested to recycle their spent catalysts. The potential applications are as follows: Catalyst manufacturers Precious metal recycling companies Electronics and lithium ion battery (LIB) manufacturers Waste management companies Modular design: reduced logistics costs and downtime Lower cost (CAPEX & OPEX) compared to existing technologies Superior recovery rate: even for low-grade wastes  Sustainable and efficient recycling: offer significant step towards decarbonisation in industrial practices Biorecycling, Platinum group metals, Low carbon emission, Decarbonisation, Clean technology, Circular economy Chemicals, Catalysts, Environment, Clean Air & Water, Biological & Chemical Treatment, Waste Management & Recycling, Industrial Waste Management, Sustainability, Circular Economy

With the global rise in ageing populations and a declining birth rate, the healthcare sector is increasingly strained, particularly in the provision of therapeutic and rehabilitative services. This humanoid technology has been designed to augment the capabilities of healthcare professionals, especially therapists, by automating repetitive intervention tasks. The humanoid enables frequent and consistent client engagement, thereby promoting inclusivity and solidarity in care settings. It works collaboratively with professionals to deliver curated intervention modules with clinically evident outcomes. By relieving caregivers of routine duties, the humanoid supports a more person-centric model of care and enhances operational efficiency in eldercare environments. Ideal collaboration partners include: Care providers – Community hospitals, eldercare centres, and day-care operators interested in deploying humanoid solutions for therapeutic engagement. Therapy professionals – Clinicians and therapists seeking technology to support non-pharmacological intervention programs. Deep-tech companies – Firms with expertise in data, image, or video analytics for enhancing humanoid capabilities. Institutes of Higher Learning (IHLs) – Academic institutions conducting research in cognitive health, psychosocial screening, or elderly care innovation. The technology combines in-house capabilities in robotic hardware and application/system software development: Hardware: Realistic humanoid form with facial expressiveness and full-body articulation enabled by 50 degrees of freedom. Software: Includes social dialogue engines, animation control, application development, and system integration. Advanced features such as video analytics and IoT edge capabilities can be developed based on partner requirements. Cognitive and physical intervention programs for elderly individuals, including dementia care. Social humanoids for community engagement, outreach, and therapeutic feedback collection. Custom humanoid deployments for enterprise-level media communication and interactive content delivery. Eldercare-focused – Designed for cognitive and therapeutic engagement in ageing populations. Deployment-ready – Proven functionality in real-world care settings. Customisable content – Delivers tailored intervention modules. Care support – Automates routine tasks to free up caregivers. Scalable platform – Built for ongoing enhancement and integration. Humanoid, Robot, LLM, Dementia Care, Eldercare, Dementia, Cognitive Health Sustainability, Sustainable Living

The world faces a mounting challenge in feeding a growing population projected to reach 9.7 billion by 2050 (United Nations). This increase drives demand for high-quality protein, but traditional sources like livestock, poultry, and fish are resource-intensive (e.g., water, land, feed), environmentally harmful (GHG emissions, deforestation) and increasingly unsustainable. With high efficiency, low emissions, and strong nutritional value, insect protein offers a sustainable alternative to conventional meat sources—especially relevant in urbanized, climate-conscious societies seeking innovation in food systems like Singapore. Crickets possess subtle flavours reminiscent of crustaceans, making them an excellent addition to our fried crackers. This familiar taste profile is particularly advantageous in Southeast Asia, where prawn crackers (Keropok) are a beloved snack. By leveraging this familiarity, this technology hopes to achieve greater consumer acceptance and rapid market adoption. These versatile crackers can be savoured as a delightful snack or paired with traditional dishes such as Nasi Lemak. Whether enjoyed as a standalone treat or as an accompaniment to a meal, these cricket-infused fried crackers offer a unique and flavourful experience that bridges the gap between innovative food trends and cultural culinary traditions. The method of processing leverages the equipment available and suitable for all standard commercial kitchens e.g. steams, dehydrators, mixers and fryers, thus allowing for lower set-up costs and being scalable to large production quantities. In addition, the recipe does not use any specialised ingredients such as modified starches, additives, preservatives. Starches used are mostly native which means the cost generally be lower and easier to source for. This makes for a relatively clean-label product. The production steps are shown below: Mixing of ingredients Precooking and drying of mixture The dried pieces are deep-fried in hot oil until crispy and golden brown This product is intended to be a high protein snack, with protein content estimated to be around 12%. It also does not contain trans fats. Sodium content can be adjusted with formulation. This makes it a healthier alternative to conventional snacks like potato chips. The shelf life of this product is estimated to be at least 6 months in proper packaging under ambient and higher if nitrogen flushed. This Cricket Keropok serves as a versatile base snack that can be customized with various ingredients, seasonings, and flavours to cater to different taste preferences and market demands. Flavour Variations with Seasonings & Spices (e.g. Mala / Seaweed) Dipping & Pairing Options (e.g. Served as a dipping snack with sauces like sambal, garlic aioli, or yoghurt-based dips) Functional & Health-Oriented Applications (e.g. High-Protein Snack – Marketed as a nutritious, protein-rich alternative to regular crackers) Innovative Culinary Uses (e.g. Crushed as a topping for salads or soups) Scalable with common kitchen equipment Clean label and free of additives Healthier choice snack Sustainable & eco-friendly protein source Customisable & versatile to cater to diverse consumer preferences Alternative Protein Source, High Protein Snack, Food Sustainability, Circular Economy, Eco-Conscious Eating, Sustainable Living Foods, Ingredients, Sustainability, Circular Economy

Monitoring safety and productivity on industrial sites is traditionally manual, error-prone, and resource-intensive. Supervisors often struggle to monitor multiple CCTV feeds, leading to missed incidents and project delays. This technology leverages AI-powered video analytics to automate the detection of safety violations—such as missing PPE, high-risk behavior, and productivity lapses—without the need for constant human oversight. In Singapore alone, over 3,000 construction-related injuries and 17 fatalities were reported in 2023, underscoring the need for smarter solutions. Beyond real-time alerts, the system delivers actionable insights to support long-term safety improvements and operational efficiency. The technology owner is seeking system integrators and software companies for R&D collaboration and test-bedding. This technology is hardware agnostic and is compatible with any IP camera or network video recorder to retrieve and analyze the video feed in real-time and provide alerts that can be sent to various messaging platforms. A server is deployed to provide the full spectrum of services such as running the software, triggering alerts, as well as the dashboard. This technology is enabled by the large construction datasets that powers object detection and tracking. The current range of detection includes scenarios such as barricade removal, workers working at height or under lifted load, safe distancing, and presence of workers in high-risk zones, PPE and more. Besides the detection of high-risk scenarios, this technology can also track productivity insights such as construction floor progress or precast lifting times. Deployment for existing use-cases can typically be completed within 1 to 3 weeks, allowing for quick integration and value realization. For newer or customized applications, the deployment timeline may vary depending on the complexity of the detection requirements and site-specific conditions. This technology can be applied across multiple industries, offering both safety monitoring and advanced analytics capabilities Construction Detection of missing PPE, unsafe behavior, and high-risk activities Time-lapse services for project progress tracking and reporting Manufacturing Monitoring worker compliance and detecting workflow bottlenecks Enhancing factory floor safety with real-time alerts Maritime & Port Operations Safety surveillance in dockyards and cargo handling zones Monitoring restricted area breaches and operational hazards Oil & Gas Detecting proximity to hazardous zones and PPE compliance Supporting incident analysis in high-risk environments Smart Cities & Facility Management License plate recognition for access control Detection of illegal parking, speeding, and vehicle trespass Medium to large construction projects are often delayed and experience cost overruns, which can be significantly improved through significant productivity gains, cost savings and early risk identification just by enabling end users to have a better understanding of their operations wherever they are which would make this a very attractive solution. Significantly improve safety hazard detection and compliance with automatic 24/7 monitoring Increase in productivity by reducing manual site inspections of up to 50% Early identification of risks to plan for mitigation Reduce human errors and ensure consistency   safety, AI, Analytics, construction Infocomm, Video/Image Analysis & Computer Vision, Big Data, Data Analytics, Data Mining & Data Visualisation, Artificial Intelligence

This non-invasive wearable integrates advanced photoplethysmography (PPG) sensing with a proprietary Pulse Shape Variability (PSV) algorithm to deliver real-time insights into stress levels linked to blood pressure fluctuations. Unlike conventional wearables that rely solely on heart rate or HRV, this technology analyzes the full morphology of the pulse waveform, capturing dynamic changes in amplitude, rise time, and contour that reflect vascular tone modulation caused by psychological stress. The result is a highly responsive and motion-tolerant stress detection platform that functions effectively in real-world conditions. By transforming microvascular signals into actionable insights, the solution enables proactive stress awareness, personalized wellness coaching, and context-aware emotional feedback, unlocking new opportunities in digital health, telemedicine, fitness, and mental wellness ecosystems. The technology owner is primarily seeking industry adopters and solution partners including medical institutions, device manufacturers, software developers, and fitness centers, who can integrate the technology into real-world applications with interest in deploying the system for use cases such as mental wellness, stress monitoring, fitness optimization, and remote healthcare. They also welcome collaboration with subject-matter experts to jointly enhance the algorithm and explore new features or application areas. This wrist-worn wearable captures and processes high-resolution biometric signals using integrated optical and motion sensors, enabling detailed physiological monitoring through advanced signal analysis. Signal Acquisition: Photoplethysmography (PPG) and Accelerometer (ACC) signals Derived Metrics: Heart Rate (HR) Blood Oxygen Saturation (SpO₂) Pulse Shape Variability (PSV) – derived from pulse waveform morphology to assess stress-related vascular responses Sleep Parameters: REM and NREM sleep stages Total Sleep Time (TST) Apnea-Hypopnea Index (AHI)  Activity Data: Step count and movement recognition System Capabilities: Real-Time Stress Detection: Continuous analysis of physiological stress signals with no need for user calibration Operates effectively during both rest and typical daily movement Motion-Tolerant Signal Processing: Proprietary algorithms reduce noise from physical activity, enabling reliable readings in dynamic conditions This wearable stress and health monitoring technology has broad applications across healthcare, wellness, fitness, and cognitive performance domains. Its ability to deliver continuous, non-invasive physiological insights makes it suitable for a wide range of use cases: Healthcare & Telehealth: Continuous patient monitoring, early detection of stress-linked health risks, and remote management of chronic conditions, particularly for mental health, cardiovascular, and sleep-related concerns. Medical Concierge & Premium Wellness Services: Personalized health monitoring for high-net-worth individuals, offering tailored stress and wellness insights, real-time biometric updates, and proactive intervention strategies. Mental Wellness & Stress Counselling: Real-time monitoring of stress indicators to support therapists, coaches, or counselors in delivering timely, personalized stress management interventions. Fitness & Recovery Optimization: Accurate tracking of heart rate and stress levels during and post workouts, enabling intelligent recommendations for training intensity, rest periods, and recovery quality. Workplace Well-being & Performance: Monitor cognitive load and emotional strain in high-performance environments, enabling preventive strategies for burnout and stress-related fatigue. Smart Devices & Platform Integration: Embedding into smartwatches, fitness trackers, or medical-grade wearables, with seamless connectivity to digital health apps, dashboards, and remote care platforms. This wrist-worn device offers a significant advancement over current health monitoring solutions by leveraging advanced photoplethysmography (PPG) technology combined with a proprietary Pulse Shape Variability (PSV) algorithm. It delivers highly accurate and continuous tracking of vital signs, including heart rate, SpO₂, and stress-related biomarkers—with minimal interference from physical movement, making it ideal for real-world, everyday use. Unlike conventional wearables that rely on basic HR or HRV metrics, this solution analyzes the full morphology of the pulse waveform to detect subtle changes in vascular tone associated with psychological stress. This allows users to correlate emotional states with verbal expressions and behavior, enabling more mindful, data-driven self-awareness and health management. The device’s motion-tolerant design, real-time data transmission, and non-invasive operation ensure consistent performance even during physical activity. Its seamless integration with health platforms and apps further enhances usability, positioning it as a versatile tool for individuals, clinicians, and wellness providers. Ultimately, this technology empowers users to make informed decisions about their stress levels, recovery, and overall well-being—bridging the gap between biometric sensing and emotional health insight in a user-friendly wearable format. Photoplethysmography, PPG, Pulse Shape Variability, PSV, Stress Detection, Wearable, Wellness, Non-Invasive Monitoring Healthcare, Diagnostics, Infocomm, Wearable Technology

The adoption of multi-functional autonomous robots is steadily increasing to support and enhance operational efficiency in the facilities management sector. This technology presents a robot integrated with advanced sensor systems, artificial intelligence (AI), and autonomous mobility to perform multiple tasks. As a digital concierge, the robot provides enhanced visitor experience with seamless 2-way communication and an integrated touchscreen to connect with site duty personnel. The same screen can double up as an announcement board for advertisements and alerts, thereby extending a virtual front-desk capability effectively. In the security domain, this robot conducts autonomous patrols with real-time video surveillance and A.I.-based anomaly detection. The security head is embedded with a "brain” to perform on-edge computing to detect security-related used cases, significantly improving safety and accuracy in complex indoor environments. For cleaning, the robot can detect over 30 types of waste with 99% accuracy. Its self-adaptive cleaning system adjusts to floor type and debris volume, while a verification mechanism ensures more effective spot-cleaning compared to conventional single-pass robots.  These Multi-Functional Autonomous Facility Management Robot can yield significant operational savings, increase patrol frequency and shorten response time to incidents. This technology offers a software that features plug-and-play solutions that be customised to specific SOPs and needs. The technology owner is looking for collaborators, such as building owners and integrated facility management companies, with use cases to test-bed AI models. Examples include but not limited to identification of suspicious baggage, illegal parking or stray supermarket trolleys.  The robot combines autonomous navigation with real-time AI processing. Its modular design allows for easy customisation based on operational needs. Key components include: Sensor suite featuring 32-beam 3D LiDAR, ultrasonic sensors & cameras for obstacle detection or environmental mapping AI-powered modules for object/person detection, thermal imaging, and anomaly alert Cloud-based dashboard for task assignment, remote monitoring, and analytics Interchangeable task modules for cleaning (e.g., vacuum/sweeper), security patrolling, and data capture Detect over 30 types of waste with 99% accuracy, with cleaning efficiency reaches up to 15,550 m²/hr Customisable module to fit specific applications This multi-functional robotic platform is ideally suited for deployment in environments that require a combination of cleanliness, security, and user interaction, particularly where operational efficiency and manpower optimization are key priorities. Industries and settings include: Commercial buildings: Automates cleaning, performs security patrols during and after hours, and assists visitors Healthcare facilities: Maintains hygiene, monitors for safety risks, and provides non-contact concierge functions Transportation hubs (airports, train stations): Enhances public safety and facility cleanliness at scale Retail complexes and malls: Supports shopper engagement, provides sanitation services, and detects anomalies Educational institutions and campuses: Ensures safe, clean, and welcoming environments Hospitality and mixed-use developments: Offers 24/7 concierge support, patrolling, and environment upkeep The global service robotics market is projected to grow by a CAGR of 30.25%, or $90.4 billion, from 2024 to 2028. This rapid growth will be driven by the continuing integration of advanced technologies such as IoT, A.I., and natural language processing into service robots. Technological advancements in machine learning, adaptive computing, and vision systems will also make service robots increasingly suitable for commercial tasks.  This autonomous multi-functional robot offers a comprehensive upgrade over current facility management solutions by integrating various functions in domains such as cleaning, security, and concierge into a single, intelligent body. This all-in-one solution delivers: Operational cost reduction through task consolidation across different functions, potentially cut cleaning and security manpower cost by 60-70% Faster response from sensing to action with integrated digital concierge, dashboard monitoring and real-time alerts Enhanced safety via advanced 3D spatial awareness Improved service quality without added manpower By unifying execution and intelligence across multiple domains, the robot transforms traditional building operations into efficient, autonomous workflows, bridging the gap between insight and action, delivering a more responsive, self-sufficient, and cost-effective solution for modern facility operations. Autonomous robotics, Integrated facility operations, cleaning automation, security surveillance, AI, robots, multi-function Green Building, Sensor, Network, Building Control & Optimisation, Infocomm, Robotics & Automation, Ambient Intelligence & Context-Aware Computing, Environment, Clean Air & Water, Sensor, Network, Monitoring & Quality Control Systems

Manual built environment inspection suffers from multiple issues such as shortage of manpower, human error and miscommunication. To overcome these issues, there is a need for an automated and centralized inspection system capable of detecting multiple defects of interest and presenting the inspection results in an easy to access format. The technology presented uses data acquired from LiDAR and Cameras mounted on an autonomous robot to inspect building interiors and external facades. The system utilizes an AI engine and can accurately detect defects such as cracks, holes, and other built imperfections stated in building quality guidelines such as CONQUAS. Defect reports can be autonomously generated after the acquired image and LiDAR data has been processed by the AI analytics engine.  The system is composed of an autonomous robot with a mounted camera and LiDAR and has the following features - Support for multiple hardware platforms such as wheeled robots and drones to allow the use of most suitable means of inspection. AI based defect detection for cracks, holes, stains, cornerness, and other structural and visual defects at >3mm unevenness. Capability to inspect for air quality, hazard detection and safety monitoring (PPE Detection). Simple and intuitive user interface with customizable repots minimizing possibility of miscommunication. Generation of defect reports compliant to Singapore Building and Construction Authority recommendations for Built Environment - CONQUAS. Defect and user management system. Capability to integrate with external Building Information Management (BIM) systems and third party apps. The system can be used for digitalization and autonomous inspection of built environments. It covers both the indoor and outdoor inspection by allowing use of multiple robot platforms. With the removal of manual inspection requirements, the system helps improve the consistency and objectiveness of inspections and helps in increasing productivity. The technology is applicaple for tasks related to management of a building from construction to maintenance. By automating and centralizing the built environment inspection, the system improves productivity, significantly reduces the time required by the inspection process, and improves the safetly of the personnel involved. The solution is useful during the entire lifecycle from construction to maintainence and provides automated reports compliant to recommendations of the Singapore Building and Construction Authority. The solution can potentially cut down the time required for inspection by several hours per residential or commercial unit. Building Indoor and Facade Inspection, Built Environment Inspection, CONQUAS, CIS-7, Quality Assessment Electronics, Sensors & Instrumentation, Infocomm, Video/Image Analysis & Computer Vision, Artificial Intelligence

Vision-based Artificial Intelligence (AI) models require substantial time to train, fine-tune and deploy in production. After production, this process is still required when performance degrades and re-training on a new dataset becomes necessary; this maintenance process exists throughout the model's lifetime to ensure optimal performance. Rather than embarking on the time-consuming and painful process of collecting/acquiring data to train and tune the AI model, many organisations have turned to the use of pre-trained models to accelerate the AI model development process. This technology consists of a suite of pre-trained models that are intended to detect food, human behaviours, facial features and count people. These AI models are operable on video footage and static images obtained from cameras. Models are tuned and trained on various use-cases and are accessible via API calls or embedded within software as a Software Development Kit (SDK) library. These models can be deployed as AI as a Service on Microservices platform providing customer data protection with blockchain technology. With customer protection enhanced with blockchain technology, AI Model performance can further be enhanced to meet customer requirement.   The technology consists of a suite of pre-trained AI models that provide high accuracy (over 80%) and can be further customised to improve accuracy and adapted to different use-case scenarios. Models can be integrated in the following ways:  Installed library package embedded within software on-device/on-premise HTTP-based Application Programming Interface (API) calls with video/image data to cloud-installed library package The following are the features for various AI models: Abnormal Behaviour Recognition Continuous monitoring and detection of abnormal human behaviours e.g. fighting, loitering Event Detection Recognises a variety of subjects and events e.g. sports day, graduation, wedding, festival, Christmas, from video footage Optimised for lightweight compute capability (Intel OpenVino) Food (Fresh and Packaged) Recognition Real-time detection of fresh and packaged foods Detects abnormal fresh food or defective packaged food Classifies food types e.g. lotus, spinach, cucumber, radish etc. Optimised for low compute capability Privacy-Preserving Person Recognition Privacy preserved people detection, counting and human activity recognition Images are blurred to preserve private information that can lead to personal identification (irreversible) Optimised for lightweight edge computing Free (Empty) Space Recognition Semantic segmentation to identify empty spaces Customisable for any free-space detection scenario High accuracy in night scenes Safety Monitoring Object detection with prohibited and allowed zones (e.g. person or forklift) Detects and identifies safety risks associated with safety distances Enables audible alarm systems of abnormal situations Wellbeing and Safety Detection Automatically detects and classifies nudity images from images  Enables alerts to be delivered to parent/caregiver's device Customisable to detect new categories of inappropriate content This technology offer comprises a suite of AI models for the following applications: Abnormal Behaviour Recognition Public areas or areas where social order needs to be maintained e.g. food & beverage, entertainment establishments Event Detection Automatic creation and/or organisation of media content i.e. photo classification Automated adjustment of device hardware parameters e.g. audio, colour, brightness when displaying specific types of content e.g. sports Food (Fresh and Packaged) Recognition Food stock level detection, food inventory management Automatic detection of fresh/packaged goods within a constrained area Privacy-Preserving Person Recognition Privacy protection of visual information, in high traffic areas, without deterioration of video quality Free (Empty) Space Recognition Vehicle position localisation on roads Navigation (free-space localisation) in partial/fully self-driving automotive vehicles Identification of free storage spaces in the logistics industry Safety Monitoring Automated compliance checks Workplace safety analysis and tracking Wellbeing and Safety Detection  Parental control in browsers, smartphones or other image storage devices e.g. Network Attached Storage (NAS), Solid State Drives (SSD) AI Models were rigorously tested in the fields of different scenarios. The microservice platform where AI Model ingest the visual data streams offers a secure customer data protection and privacy using blockchain technology. Making this Microservice platform capable of tracking customer’s data usage and thus ensure privacy when AI model operating on the platform are simultaneously improved using unique customer data captured on customer’s premise. Accelerate AI development - eliminate the need for dataset creation, annotation, tuning and testing Customisable AI models - fine-tuned to environment and condition Operational support to continuously improve AI accuracy from newly collected data   event detection, abnormal human behaviour recognition, safety monitoring, food package detection, food freshness, nudity detection, empty space Infocomm, Video/Image Analysis & Computer Vision, Video/Image Processing, Artificial Intelligence