As lockdown restrictions begin to ease, changes to the transport infrastructure across Oxfordshire are under way to create space for people to get around while maintaining social distancing. Oxfordshire County Council, which runs the county’s road networks, is keen to avert a return to pre-lockdown levels of car use and the high pollution and congestion that would bring, and is turning to data-based evidence to figure out what road use might look like and what changes will bring about the best outcomes. Its new transport modelling platform MIMAS is currently in development, but may still be able to offer some good insights.
Even without Covid-19, significant changes to Oxfordshire’s transport networks are in sight, as planners prepare for more electric vehicles, connected autonomous vehicles (CAV’s), e-scooters and e-bikes on the roads. The climate emergency, declared by the council in 2019, is also adding pressure to enable motorists to leave their cars at home and make active travel choices.
Government funding for these policy shifts was announced in February this year, when £5 billion was allocated nationally for sustainable transport. This included £2 billion for cycling and walking schemes, the first £250 million of which was released in May.
When deciding how to spend its share of this money, Oxfordshire County Council is drawing on data-based evidence to inform decisions. MIMAS is to replace a seven-year old modelling platform and, while not yet operational, the preparation work being done is likely to bring new insights into how traffic flows will have changed during the lockdown. When it goes live, the model will be more adaptable and responsive to changing conditions than its predecessor and will use live transport data, refreshed monthly, combined with simulations.
Reynold Greenlaw is director of innovation delivery for Oxford Computer Consultants, the firm leading the MIMAS project. “We don’t know what the new normal for traffic flows will be, and neither does anyone else,” he said, explaining that the new model will be able to take pre-lockdown data and compare them with traffic flow data post-lockdown and see how they move. “It won’t be perfect but it will be better and more nimble than anything we have.”
Getting access to robust data on how traffic has changed during the lockdown is surprisingly hard. Most people will agree that the roads were remarkably quieter in March, April and May, and there were noticeable upticks in traffic as movement restrictions began to ease. But how many more vehicles, what times they are using the roads, how that differs in different roads or areas, is difficult for councils to measure. Even the government’s Department for Transport (DfT) was struggling to gain access to transport data, and in March put out an urgent call to local authorities for data sets on traffic counts, trip lengths, parking, cycling and walking.
Alchera Technologies specialises in real time data analysis in transport systems and is one of the partners in the MIMAS project. The firm’s co-founder Anna Jordan explained that the data required to understand traffic flow changes during lockdown is not available by default and there’s a cost to accessing it: “For example, CCTV footage won’t be stored for privacy reasons, so we will be taking data from ATC’s [automatic traffic counters, often installed as rubber tubing across roads to measure vehicle numbers] and analysing them as historical measurements.”
Alchera takes data from various live sources such as CCTV feeds and ATC’s, then uses machine learning to identify different modes of transport such as cycles, cars, buses and trucks. These can be combined with data from bus companies or fleet management systems like Teletrac Navman, which give comprehensive measurements on vehicles’ movements. Jordan describes the firm’s role as building real time data platforms to make infrastructure intelligent: “We effectively build a ‘digital layer’ on top of the physical world.”
With MIMAS, Alchera is, Jordan explained, creating a “real-time data exchange, assessing measurements from camera footage, vehicle counts and classifications, which will be fed into the model, to make it up to date and to be something that can be used in a responsive fashion. It will give us a better understanding on how people move around and provide a model that is more true to life and reflective of reality. MIMAS will take into account historical traffic patterns and movements, but we will also measure what’s happening now.”
Oxfordshire County Council’s existing transport model runs on aggregated, historical data from phone companies, acquired by the authority every few years and validated for the Oxfordshire region, a normal process in the world of transport planning.
For Jordan that time lag is no longer acceptable. “If the council wants to install a cycle lane, that will have an impact and you have to understand what that will be. You have to be able to test that iteratively, and the model’s responsiveness is key.
“To be able to respond to how society is changing, you have to let people move around. You can’t build a model that changes every five years,” she said.
Jordan’s ambition for MIMAS is to provide more real time measurement. “At some level we are gathering [real time] information, but we are not yet fully integrated to the council’s systems.”
Data analysis which Alchera has done for London during the lockdown indicates what MIMAS could deliver for Oxfordshire. “Here we are analysing data from across London including TfL (Transport for London) OpenData video cameras and other sensors, including those from Highways England. These were costly for TfL, but now provide rich and valuable data for analysis built on top of it,” said Jordan, pointing out that TfL’s data is the basis for London’s presence on the global app CityMapper.
“We were able to uncover interesting behaviour, such as different responses to lockdown in different geographical areas,” she added, such as car traffic in central London dropping less after the Covid-19 restrictions (15%) than it dropped in the peripheries (42%). “As soon as you see that, you have your finger on the pulse on how behaviours are changing. So after a Boris Johnson announcement, you can see how it will change the composition of traffic the following day. To have that real time understanding will help when it comes to crisis and policy response.”
Alchera Technologies was established in 2016 with the aim of fusing together machine learning with the transport infrastructure sector and, as Jordan explained, to “build something that works better for citizens. There’s nothing that impacts people and how they live their lives more than transport. We’re moving into a world which is more crowded, and journeys are more challenging than ever.”
Laura Peacock, innovation hub manager for Oxfordshire County Council, explained that to help the region to emerge from the lockdown, the team is helping operators and policy makers with the evidence and data they need on mobility behaviour, air quality and public health to make effective informed decisions.
Peacock recognises MIMAS as the first modelling approach a council has used that offers a user-centric, agent based, perspective instead of a systems-based approach.
This offers potential for the model to provide the evidence-based understanding of why, for example, parents would choose the car over the bus when taking their children to school.
The council is drawing on the experience and knowledge it has gained in mobility modelling in recent years. “This,” she said, “will give a useful understanding of how people are socially distancing and where the hot spots are in public spaces.
“The centre of Oxford is medieval, with many pinch points and narrow public spaces, so our data modelling and analytics experience helps us to understand how we can re-prioritise road space, and how we avoid going back to the normal ways, making public transport and sustainable transport options the safe choices.”
Strategically Oxfordshire County Council is looking at how it can present evidence to generate or divert funds towards making positive changes to the transport network for the longer term. The innovation hub is developing a framework bringing together all the work and experience the authority has done around innovation in modelling, procurement, healthcare and future mobility, including the integration of electric and autonomous vehicles, and even drones.
Peacock, a strategist by profession, is aware that transport can be the biggest barrier to planning, and the viability of any urban plan will come down to the transport infrastructure around it. “We aim to set policy guidelines together with evidence,” she says, “to give planners and developers the confidence to push ambitions to the next level, providing future developments that are fit for purpose and flexible to change as citizens’ aspirations change. We will also collaborate with central government looking to support an innovative and flexible planning system, taking examples of best practice in Oxfordshire and around the world.”
MIMAS should be in use by the end of this year and it is hoped that other councils will adapt it too, as it considerably lowers the cost barrier for modelling. As Greenlaw summed up: “MIMAS may begin with a scenario set up by an expert, but instead of a report, the council will get a tool which they can play around with and explore different scenarios, without paying extra.”
The consortium partners developing MIMAS are agent-based modelling firm Immense Simulations, location-based big data specialist GeoSpock, and real-time journey analysis providers Alchera Technologies and Zipabout. Oxford Computer Consultants is managing the consortium and is handling user research, UX design, data visualisation and system integration.
Access Alchera Technologies’ report and Covid-19 dashboard Digging Deeper into Transport Behaviour: City-scale AI analysis to understand London’s response to Covid-19