Robok Limited Public Funding

PortFlow, led by RoboK Limited in partnership with Red Funnel and BioBright, is a six-month pre-deployment trial focused on reducing greenhouse gas emissions from short-sea ferry operations through improved operational scheduling, real-time data integration, and environmental monitoring. Ferry terminals are busy, high-traffic environments where vehicles often arrive well in advance of departure times. This frequently leads to congestion, long idle periods, and increased emissions from both vehicles and vessels. Traditional ferry operations rely on fixed schedules and manual coordination, limiting their ability to respond dynamically to changing demand or real-time conditions. PortFlow will address this challenge by developing a prototype platform that applies artificial intelligence (AI), computer vision, and emissions monitoring to optimise ferry terminal operations. The platform will use data from existing CCTV infrastructure, vehicle check-in records, and vessel activity logs to provide real-time visibility of traffic flows and vessel readiness. By monitoring vehicle arrivals, advising optimal check-in times, and detecting congestion or bottlenecks, the system will help operators to reduce dwell time --- the period that vehicles and vessels spend idle --- and therefore cut associated fuel consumption and emissions. To support accurate measurement of environmental impact, project partner BioBright will deploy emissions monitoring sensors within the terminal environment. These sensors will capture real-world carbon and air quality data, allowing the project to validate the emissions reduction potential of smarter scheduling and operational optimisation. The platform will also feature a user-friendly dashboard, providing ferry operators and port staff with live data, emissions insights, and scenario planning tools to support operational decision-making and long-term planning. PortFlow contributes to the UK's broader net zero ambitions and Clean Maritime Plan by demonstrating how digital tools, AI, and low-cost infrastructure can deliver emissions reductions in the maritime sector. The project offers a scalable, cost-effective approach that can be adopted by ferry operators and ports across the UK and internationally. By combining predictive scheduling technology with real-world emissions monitoring, PortFlow supports a more sustainable and data-driven future for ferry operations --- delivering both environmental benefits and operational efficiency improvements.

HARBOUR-AI is an innovative collaboration designed to upskill maritime professionals in using AI technologies to improve safety, operational decision-making, and environmental sustainability. Developed by RoboK Limited in partnership with proposed end users Falmouth Harbour Commissioners (FHC) and A&P Falmouth, the project addresses the challenges of managing diverse maritime activities in the Great South West region, recognised as the UK's blue capital. The HARBOUR-AI platform transforms existing CCTV infrastructure, the primary tool for monitoring port environments, into a personalised AI training and competency development solution. It equips port workers and technical teams with the skills to interpret, train, and refine AI tools, enabling them to: * **Identify Hazards**: Learn to use AI for vessel tracking, anomaly detection, and behavioural analysis to proactively manage risks and enhance port safety. * **Improve Decision-Making**: Use AI-driven traffic analytics to support Navigational Risk Assessments and optimise resource planning. * **Enhance Environmental Monitoring**: Gain skills to train AI models for detecting hazards such as oil spills, unauthorised anchoring, and emissions hotspots. By integrating role-specific training aligned with the **AI Skills for Business Competency Framework**, HARBOUR-AI ensures that port professionals develop the necessary competencies to adopt AI responsibly and ethically. This includes fostering awareness of data privacy, ethical considerations, and the limitations of AI tools. Designed to be accessible and scalable, HARBOUR-AI eliminates the need for costly hardware upgrades, making advanced AI capabilities available to small and medium-sized ports. This cost-effective solution supports workforce upskilling, sustainability goals, and operational improvements, all while enabling ports to become leaders in responsible AI adoption. HARBOUR-AI is a transformative step toward smarter, safer, and greener port operations, empowering the maritime workforce to harness the full potential of AI technologies.

We introduce **PALLETS**, **P**roactive **A**I-powered so**L**utions for **L**ogistics **E**fficiency, **T**ransparency and **S**afety. It is a holistic AI-based analytics platform tailored for the logistics and transport sector to improve operations and address existing bottlenecks that hinder the broader adoption of AI and ML in the field. **Highlights:** 1. **Democratisation of AI in Logistics**: The PALLETS platform is designed to increase transparency and bring secure AI-driven computer vision solutions to stakeholders from different parts of the supply chain, enhancing visibility at its various nodes utilising CCTV cameras 2. **Use cases for SME and large organisations:** * _**Safety**_ **-** Hazard Detection: Identifying and mitigating potentially hazardous events to enhance safety. * _**Efficiency-**_Dwell Time Tracking: Precise measurement of goods or vehicle stay duration for improved operational efficiency. 3\. **Addressing AI Adoption Challenges:** the solution seeks to: * Ease Data and Performance Concerns: By streamlining the ML development cycle from data to model deployment to performance monitoring in production, we automate manual processes and reduce data required for training and developing and also enable faster iterations and continuous improvement. * Enhance Trustworthiness: We introduce privacy and security by design from the onset of solution development - ensuring adoption of best practice, identification and mitigation of potential vulnerabilities in the platform, improving end user trust related to data and AI solution's security and reliability. 4\.**Human-Centric Solution for Dynamic Environment**: designed for user experience for individuals responsible for operations in highly active areas such as ports, warehouses and yards, PALLETS is designed with human centricity at heart to be assistive and empowering, providing relevant and useful information for multitasking personnel in these bustling locations. 5\. **Innovative Technical Architecture**: * Automated ML Pipeline and Analytics Engine: reduces initial on-boarding processes for customers with varying sites, cameras and unique needs for scalability with automated, continuous improvement to achieve solution reliability and cost-effectiveness. * Privacy and security by design: incorporate key security and privacy principles and best practices, addressing vulnerabilities throughout. * Human-in-the-loop Integration: allows user input and actions as part of work processes. * Active Learning & Feedback Loop: AI-assisted data curation for reduced data and faster development. **Potential Impact**: The "PALLETS" project promises enhanced logistics visibility and efficiency, benefiting SMEs and major organisations in the supply chain industry. By addressing AI adoption hurdles and emphasising safety, efficiency, and transparency, the project has the potential to revolutionise logistics and transport operations and safety standards.

Our project, "SeeGull: AI Tools for Enhanced Logistics Productivity," utilises cutting-edge artificial intelligence technology to transform transport and logistics resource monitoring and allocation. By deploying AI algorithms and strategically placing cameras throughout logistics facilities, our system aims to improve logistics efficiency through increased transparency and observability. **Key Innovations:** * **Activity Analysis**: SeeGull's AI system continuously analyses visual data from multiple camera feeds to detect movements of forklifts, pallets, and lorries. This data helps identify operational bottlenecks and provides valuable insights for performance optimisation. * **Anomaly Detection**: Using advanced computer vision algorithms, our AI can identify anomalies within the warehouse, such as unexpected obstructions or prolonged dwelling. This ensures the maintenance of safety protocols and the smooth running of operations, reducing costly errors and accidents. * **Resource Allocation**: SeeGull's system predicts and plans resource allocation, including forklifts, personnel, and storage space, by utilising historical and real-time data. This dynamic resource allocation enhances efficiency, minimises bottlenecks, and optimises overall logistics performance. * **Scalability**: Our platform is designed to seamlessly expand as more warehouses and storage areas are covered. The addition of more cameras enhances the system's intelligence and efficiency, making it a valuable investment for businesses of all sizes. * **Integration and User-Friendly Interface**: Recognising the importance of ease of use in the logistics industry, our design focuses on integrating seamlessly into the daily workflow of facility managers and operators. It can be integrated with existing warehouse management systems to provide a unified source of information, simplifying the transition to a smarter, safer, and more efficient logistics operation. * **Collaboration and Visibility**: We incorporate collaborative features that facilitate communication and transparency, allowing users to log and share actions and activities, ensuring enduring visibility throughout the logistics process. * **Compliance and Audit Trail**: SeeGull provides comprehensive information logging, simplifying safety and regulatory compliance for businesses. This feature streamlines the reporting and auditing process, reducing administrative burdens. In summary, our project leads the way in logistics innovation, combining AI, computer vision, and predictive analytics to enhance safety and efficiency. By delivering real-time operational insights, anomaly detection, and adaptive resource allocation, our solution empowers logistics companies to optimise their operations, reduce risks, and achieve greater transparency and observability throughout their facilities. This not only improves safety but also enhances overall efficiency, providing businesses in the logistics industry with a significant competitive advantage.

We introduce INTERMODAL, INTElligent Real-time, MOnitoring & Detection video AnaLytics, an innovative AI-driven solution designed to enhance security, safety and operational efficiency at large-scale rail construction sites through real-time video analytics, with the potential to benefit the wider infrastructure sectors. **Highlights**: **1\. Challenges in the railway construction:** * **Safety and surveillance concerns**: Inefficiency & safety issues cost £1.4bn/year to UK economy with construction workplace injury and ill health(HSE, 2019/20). Security issue like theft and vandalism cost is estimated £800m/year. Over £2.2bn/year economic value (2% of output) can be potentially recovered with improved safety & security at construction sites in UK alone. * **Passive and reactive processes**: Conventional approaches heavily rely on manual oversight and passive surveillance, often reacting to incidents rather than proactively implementing measures to prevent them. * **Lack of timely and structured informatio**n: Obtaining timely and structured information for managing large, dynamic industrial sites can be challenging and can lead to various operational efficiency issues. **2\. Innovative approach** * **Advanced security and operational monitoring**: our proposed solution goes beyond traditional manual monitoring for security. Instead, AI analytics can monitor movements of personnel, equipment dynamics, and behavioural analysis within the railway construction site. * **AI-powered analytics**: The live videos from CCTVs are actively processed for identifying potential risks, anomalies, or inefficiencies. Whether it's predicting possible human-machine near misses or detecting perimeter breaches, the AI acts as an ever-vigilant sentinel, offering real-time insights to site managers. * **Robust tracking algorithms**: built for both indoor and outdoor, busy or cluttered environments, ensuring accuracy of behavioural monitoring in various conditions. * **Human-in-the-loop Integration**: allows user input and actions as part of work processes. * Continuous improvements: self-learning nature of our AI means the system continually evolves, adapting to the unique dynamics of each railway site and offering more refined insights over time. **3\. Use cases for GCRE and beyond construction phase** * Security-Proactive surveillance: Enabling real-time, actionable insights into live events * Safety-Hazard detection & mitigation: Proactively identify and mitigate potential hazardous events, ensuring increased safety within construction environments. * Operational Efficiency- Dwell time tracking: Accurate measurement of where and how long goods, assets or vehicles stay in one place, enabling improved operational efficiency. **4\. Potential impacts** Our solution revolutionises surveillance by replacing traditional, labour-intensive CCTV monitoring with dynamic, real-time AI-driven intelligence. Through advanced monitoring and predictive analytics, we enhance the resilience of railway construction projects. Early risk detection and proactive mitigation minimise disruptions and maximise success for infrastructure companies

We propose CHECK, an AI-powered solution that uses federated learning and computer vision to build a privacy-preserving, scalable analytics platform for small and medium-sized Third-Party Logistics Providers. The proposed deployment of CHECK will run on a distributed architecture, connecting multiple physical sites and nodes in the logistics process from inbound transport to warehouse/storage to dispatch. The goal of the solution is to help create visibility and trust of how AI can be used to securely share insights re. capacity, resource availability and worker/vehicle interface to bring safety and efficiencies in their operational processes. Small logistics companies are faced with a number of challenges including limited resources, technologies and capacities - potentially hindering their competitiveness among global, larger 3PL (third-party logistics) companies. Seeking to help address these challenges, CHECK intends to build a consortium for a project that will turn distributed, passive videos into real-time, always-on, proactive analytics engine that would provide efficiency and safety benefits for logistics operators specifically in the transportation & warehousing processes. The goal is to to ensure smaller logistics providers' market competitiveness, keeping their costs down, operations efficient and workplaces safe.

Due to the nature of the work, construction is one of the most high-risk industries in the UK. There were 30 fatal injuries to workers in 2021, with top safety hazards incl. working at heights, struck by moving objects, slips, trips & falls, and material handling. Additionally, security issues such as theft and vandalism present not only direct replacement costs but also potential delays to project delivery. It is estimated that safety & security issues cost UK construction industry £2.2bn per year. Project INTERMODAL - INTElligent Real-time, MOnitoring & Detection video AnaLytics solution for rail construction, seeks to introduce a highly intelligent, real-time, automated video analytics solution to improve security and safety for the construction of a very large-scale rail site, with the potential of extending such solutions to the daily operation and maintenance beyond construction. In particular, INTERMODAL turns passive surveillance videos into real-time, always-on, proactive analytics engine that would detect & monitor for a wide range of key use cases aiming at addressing key security & safety operation challenges for the construction phase of GCRE.

Our vision for the future of UK infrastructure encompasses fleets of unmanned aircraft systems (UAS) powered by renewable energy to deliver more efficient applications and processes in high cost areas such as search and rescue and highway maintenance, whilst reducing the economic impact currently caused by interruptions such as road closures. Additionally, using UAS in such operations can significantly reduce their carbon footprint. This vision will be delivered by 2025 at scale. The InDePTH project will investigate the use of autonomous drones to deliver this vision. The aircraft will be used to regularly survey wide infrastructure estates, including ports and highways, to create digital models and obtain detailed insight of these dynamic environments. InDePTH will utilise onboard sensing, data and image processing equipment to autonomous drones, currently available as drone-in-a-box (DIAB) solutions. Current DIAB offerings include mission-tailored sensing equipment and minimal human input and supervision but lack end-to-end and real-time data analytics integration. DIAB solutions today require lengthy manual data offloading after missions, making real-time analytics impossible. Another constraint of current DIAB solutions is that data offloading is typically not fully integrated with analytics software, requiring the use of cloud-based data lakes. The project aims at fast-tracking data transport while providing enhanced AI analytics near real-time. InDePTH will augment the drone data analytics using state-of-the-art machine learning (ML) algorithms developed by RoboK, creating optimised image processing aiming at modelling environments to a 3D digital twin. BT will provide secure and fast data transport equipment by exploring the use of fast and reliable 5G and fibre links to transmit DIAB data with low latency. Three demonstrators will be developed to support critical use cases for Associated British Ports (ABP) and Kier Highways. Port and highway environments change rapidly due to constant movement of people ,vehicles and goods. For ports, two key use cases are identified: InDePTH will investigate the use of UAS to improve inventory management for ABP ports, focusing on vehicle inventory; furthermore, ABP will use drones in their off-shore surveillance and maritime operations in the second project demonstrator. Thirdly, in the highways area, InDePTH will look at deploying UAS to continuously assess the ground surface quality of highways, for Kier.