InstaDeep explores the benefits of Deep Reinforcement Learning technology in transportation at Interchange event



Earlier this month, InstaDeep attended Interchange, an event that brought together a diverse group of delegates, from infrastructure owner-operators, local authorities, system operators and specialists in the transportation industry. The event offered a comprehensive programme, providing a wealth of opportunities for attendees to gain insights into various transportation-related topics. 

During the event, InstaDeep – a Google Partner Advantage Service and Build partner – presented a talk as part of the Google Cloud Prism programme. The programme sought to explain how data, advanced technologies, and innovations are reshaping the UK transport network, enhancing its performance, sustainability, and user experience. By leveraging the latest cutting-edge solutions, it strived to promote a more integrated and streamlined transport system that delivers optimal outcomes for all stakeholders involved.

Shared synergies 

InstaDeep’s expertise in AI and digital innovation, along with its proven ability to deliver effective real-world applications, closely aligns with the objectives of Interchange. The event was focused on integrated and intelligent transport infrastructure, bringing together a new community of infrastructure operators, designers, builders, place makers, and technologists to improve places and deliver net zero. It envisioned a new era of collaboration and joined-up thinking, where traditionally separated transport disciplines such as road, rail, air, marine, and active travel are reconciled with the real estate, digital, and energy communities. It covered a range of topics, from passenger and user experience, to leveraging data for operational efficiency and addressing cyber-security. Attendees gained insights into how historical location data can be used to inform transport decisions today and plan for the future, while also exploring the strategies for delivering optimal digital experiences to passengers.

Leveraging advanced AI for real-world problems

During his session, Dr. Waqaas Rehman, one of the company’s Business Development Leads and a logistics and transportation domain expert, demonstrated how Deep Reinforcement Learning can be leveraged to reimagine capacity planning and traffic management. 

By way of example, he explained how InstaDeep is partnering with German operator Deutsche Bahn, to revolutionise the country’s national rail system as part of their Digitale Schiene Deutschland initiative. By leveraging their deep expertise in AI, the InstaDeep team aims to increase network capacity and reduce delays in real-time and in a manner that finds the optimal outcome. As Waqaas commented, higher train operating frequency is only possible if the infrastructure side can provide and manage the required capacity and ensure a seamless traffic flow for operators. To further support this perspective, he presented several other transportation case studies where insufficient network capacity and service delays were identified as significant issues. He then explained the company’s approach to dealing with these limitations, pointing out that organising traffic in a railway system has always needed a lot of planning. In the digital age, companies can turn this into an advantage: if they can get all relevant data into one place, and in real-time, they will hold the key to orchestrating the entire operation in ways that the present, human-centred processes do not permit.

Through the partnership with Deutsche Bahn, InstaDeep is making significant strides in developing an AI-powered capacity and traffic management system (CTMS) that can address real-world problems, from automated timetable planning and scheduling of missions and possessions (e.g. when a section of the track is blocked for maintenance work), to optimised re-dispatching, while maintaining interoperability with other operators in the EU. Waqaas noted that InstaDeep has been working on it since 2019, leveraging recent breakthroughs in AI to replace traditional planning and control systems. Through its advanced capabilities, the CTMS will improve service quality, enhance capacity, and optimise the use of resources such as staff, material and energy. As such, it represents a significant step forward in the modernization and digitalization of railway infrastructure. Illustrating the system’s capabilities, he explained how it will soon “automatically make timetables and real-time dispatching decisions for more than 10,000 trains per day on a network exceeding 33,000 km.”

Looking to the future

Waqaas also outlined the company’s future plans to expand its AI research and capitalise on its expertise to create customised AI-powered solutions for its other major partners in other industries. He drew attention to the recent announcement by BioNTech, the global biotechnology company, to acquire InstaDeep, which has the stated aim of building world-leading capabilities in AI-driven drug discovery and the development of next-generation immunotherapies and vaccines. 

Summarising the Interchange event, he commented: “Interchange brought together like-minded people keen to learn and share knowledge about the latest advancements in AI technology. So many experts and innovators came together to discuss the potential of AI in revolutionising the transportation and logistics industries. It was inspiring to see how the transportation industry is leveraging AI to solve complex problems, and we are proud to be at the forefront of this innovation. 

>> If you are interested in joining a company whose mission is to drive innovation and advance the development of AI, InstaDeep is actively hiring for many positions worldwide. View our open job opportunities and apply today!