RoboK case study

RoboK specialises in developing efficient AI-based computer vision solutions aimed at improving safety in transportation. RoboK’s work with TfW (Transport for Wales) aimed to address level crossing safety challenges on their network.

As winners of cohort 3, RoboK collaborated with TfW to conduct a proof-of-concept (POC) phase, consisting of two stages, which demonstrated the effectiveness of their solution.

The RoboK developments within the programme lead to subsequent adoption of RoboK’s solution across the Core Valley Lines, their selection to work with the Global Centre of Rail Excellence, and their successful grant application to the Network Rail’s Performance Innovation Fund.

Phase 1: Proof-of-concept (POC) installation at Ty Glass level crossing

TfW embarked on a comprehensive safety improvement initiative by partnering with RoboK. The POC phase involved the installation of a solar-powered camera at Ty Glass level crossing, enabling continuous data collection over a period of 31 days. RoboK’s AI-based sensing technology analysed the collected data, providing valuable insights that evolved over time.

Phase 2: Refinement and analysis of AI-based sensing technology and 2: Sleeper change and fleet tasks

During the POC phase, RoboK’s AI-based solution continuously refined its capabilities. Through an iterative learning process, the technology developed a deeper understanding of the unique challenges associated with level crossing safety on TfW’s network. This resulted in the production of insightful data, which shed light on potential safety hazards and further informed the development of the solution.

Implementation on the Core Valley Lines

Based on the success of the POC phase, TfW decided to extend the use of RoboK’s solution to the Core Valley Lines. By leveraging RoboK’s AI-powered computer vision technology, TfW aimed to enhance safety measures across this significant section of their rail network. The solution’s comprehensive coverage enabled early detection and mitigation of safety risks, reducing the likelihood of accidents or incidents at level crossings.

Partnership with the Global Centre of Rail Excellence

RoboK’s expertise and success in addressing level crossing safety challenges led to their selection as a strategic partner by the Global Centre of Rail Excellence. This partnership fostered collaboration and knowledge sharing, empowering RoboK to refine and enhance their computer vision solution for the rail industry. By leveraging the expertise of the Global Centre of Rail Excellence, RoboK further bolstered the effectiveness and applicability of their solution.

Grant application and funding success

Building on the positive outcomes achieved during the POC phase and subsequent implementation on the Core Valley Lines, TfW, in collaboration with RoboK, successfully applied for funding from the Network Rail’s Performance Innovation Fund. Their grant application was approved, resulting in an award of £200,000. This funding will be utilized over the next two years to develop and customize the AI-based computer vision solution specifically for the Welsh Network. The financial support ensures the continued refinement and optimization of the solution to address the unique challenges faced by TfW. 

RoboK’s partnership with TfW has led to significant advancements in level crossing safety on the Welsh rail network. The successful completion of the two-phase POC installation at Ty Glass level crossing showcased the efficiency and effectiveness of RoboK’s AI-based computer vision solution.

The subsequent implementation across the Core Valley Lines demonstrated the scalability and wider applicability of the solution. Additionally, RoboK’s collaboration with the Global Centre of Rail Excellence and the awarded grant from the Network Rail’s Performance Innovation Fund further validate the industry’s recognition of their innovative technology. The ongoing development and customisation efforts for the Welsh Network promise to significantly enhance transportation safety and mitigate risks associated with level crossings.