Digital Twinning for Resilience in the Transport Sector
Adam Fenton
Digital Twins may just be the most important aspect of Artificial Intelligence that you’ve never heard of, yet. As the name hints at, a Digital Twin or DT for short, is a digital replica of a physical thing. It could be anything from an aircraft to a bridge or even a more complex system like a hospital or a road network. The crucial key is that – distinguishing a DT from a mere digital ‘model’ or ‘shadow’ – there is a bidirectional data flow between the DT and the Physical Twin (PT). The physical version sends data to the DT, which can be analysed using algorithms, machine learning or neural nets, and is then linked back to the physical twin. In this way the physical world affects the digital twin, and the DT in turn can make changes and affect the physical twin and hence the physical world. This process can be repeated ad infinitum and the DT can be used to optimise the efficiency or performance of the PT.
An interesting and illustrative example of this is Google Maps. The map is the digital version of the physical road network. By receiving data from cars on the road, the map is able to model traffic build-ups on various routes, and using AI is able to suggest alternative routes to approaching cars, to alleviate traffic on that route. Hence the DT receives data from the real world, and uses that data to affect conditions in the real world in a continuous loop. By this definition, Google Maps fulfils the criteria for a digital twin.
With its roots in NASA’s Apollo space program of the 1970s, DTs have been used in industrial and engineering processes for decades. A DT of an aircraft for example is a unique twin of an individual aircraft. Throughout its product lifecycle, data is continuously fed back and forth between the DT and the PT from an array of sensors throughout the aircraft providing a stream of continuous data for monitoring and processing. The DT can provide warnings if any part of the PT is showing signs of stress or abnormality, and hence requires attention or maintenance.
Recent advances in Internet of Things (IoT) devices, sensors, 4G and 5G network coverage, AI, Machine Learning, and Distributed Ledger Technology, mean that the potential of DTs to be applied in areas beyond manufacturing and engineering, is being intensively explored. The UK government is heavily invested in the exploration of DTs for other uses and established the National Digital Twin Program (NDTP) a “government-led programme committed to growing national capability in digital twinning technologies and processes throughout the country”. As the UK Government Office for Science points out in its Rapid Technology Assessment (RTA) of DTs “as the technology develops digital twins could present benefits to sectors beyond industry and manufacturing, including health and defence”.
The UK Department for Transport (DfT) is similarly investing in exploringthe many potential benefits of DTs in the transport sector with a dedicated Digital Twins program within its Advanced Analytics Division. The Transport and Research Innovation Board (TRIB) has developed a vision and roadmap “to enable a trusted ecosystem of connected digital twins for multi-modal UK transport networks” by 2035. Hence, along with investigating efficiency gains in decarbonisation and day-to-day operations, the DfT is funding cutting-edge research into the possible applications of DTs in building resilience to disruption within the transport sector. Funded via UKRI’s Engineering and Physical Sciences Research Council (EPSRC) the Digital twinning for crisis response in transport-based scenarios project is a collaboration between the Digital Twinning Network + (DTNet+) and Resilience Beyond Observed Capabilities Network + (RBOC N+), with lead researchers from the Alan Turing Institute and Coventry University.
DTBOC (Digital Twins Beyond Observed Capabilities) currently underway, will conduct two Sandpit events bringing together over 100 academic researchers, industry and government representatives to explore the uses of DTs for crisis response in the event of disruptive events, both natural and human-caused, such as pandemic, flooding, terrorist or cyber-attack, rioting, strike and accidents. It will also fund 3-5 further proof-of-concept research projects focussed on ideas generated during the sandpits. It is clear, there is huge potential for DTs to play a part in optimising preparation, response and recovery from such destructive and disruptive events as those mentioned above. DTs could provide advanced warning based on predictions and forecasts generated from combining historical and current weather data for example.
Beyond just predictions though, DTs, combined with AI could provide recommendations for policy interventions, decision support, route optimisation and resource allocation support in catastrophic emergencies. The DT could for example automate processes for optimising the allocation of emergency vehicles in a scenario with multiple points of crisis, matching the available resources, equipment and staff to the needs on the ground. DTs using data from diverse sources such as social media, sensors, cameras, drones, and other devices could provide an overall, real-time situational awareness for decision makers.
This could provide first responders, across multiple agencies, with real-time data on the numbers and locations of victims and people in need of help in the aftermath of a terrorist attack or catastrophic accident. In short, DTs could help to save lives during emergencies. There is much interest in the use of modular or federated DTs which would be comprised of multiple interconnected systems, or as the RTA states “connecting digital twins can increase their value and power, creating a ‘system of systems’. This could support improved decision-making in complex areas and form the basis of a federated UK Digital Twin”.
As a key part of the transport sector, maritime applications are included in the review of the entire transport network. A recently published report commissioned by DfT notes that DTs for freight management at ports could reduce emissions by decreasing turnaround times for cargo ships and result in millions of pounds in efficiencies. Research by DTBOC suggests that DTs could be used to monitor the network of container freight coming into the UK and be used to provide risk analysis, early warnings and recommendations for dangerous or illegal cargo.
The potential links to the National Maritime Single Window, a rich source of data for maritime-related DT applications, should be explored in more depth. DTs could be used to monitor remote networks like subsea cables and pipelines for abnormal activity. Other applications would boost the efficiency and resilience of the maritime sector as an integral part of connected transport networks, for example boosting cybersecurity by monitoring cyber trafficacross networks, and building resilient supply chains resistant to disruptions like pandemics. Virtual and Augmented Reality applications for training emergency responders is another potential area of great interest for responding to fire, earthquake, explosions, attacks or accidents. One scenario DTBOC participants were asked to consider was a cyber-attack on a ship causing a collision and closure of a major UK port.
As always in debates around AI there is potential for enormous benefits but a corresponding range of challenges and risks. Data sharing and interoperability across networks, departments and sectors have been identified as key ethical, societal and legal challenges. Ensuring that the benefits of technological advances flow to society and not into the pockets of Big Tech is another. Initiatives such as this one from DfT show an intention to grow a base of expertise and knowledge within the UK, and to ensure that the benefits accrue to the public in the form of more efficient and safer transportation networks.
For more information about DTBOC see the project page at: https://rboc.ac.uk/pages/dtboc