Engineering startups vs Covid-19: How can AI technology help the UK’s Covid-19 response?

31 Jul 2020

Vivacity labs banner

The nationwide implementation of social distancing rules has been one of the UK’s most pressing challenges throughout the pandemic and the easing of lockdown. An evidence-based and data-driven approach has been key to making broad-ranging decisions in a timely manner and the government has turned to engineering and technology startups for help. In this blog, we explore how Hub Members Vivacity Labs, Slingshot Simulations and Humanising Autonomy have adapted their AI-based innovations to inform the government’s Covid-19 response.

Vivacity Labs

Who are they?

Vivacity Labs have patented machine-learning technology that continuously captures and classifies real-time pedestrian and transport usage in urban areas. Using a thousand AI-programmed cameras and sensors in a dozen towns and cities, the company enables local and regional governments to analyse, forecast and improve traffic flows, networks and air quality. They do this on a micro level by seeing how different types of vehicles interact and on a macro level by surveying activity at city-scale.

In 2016, co-founder Yang Lu joined the Enterprise Fellowships programme to take their small business to the next level with the Hub’s financial backing and business support. Since then Vivacity Labs have gone from strength to strength, by securing a partnership with Transport for Greater Manchester to control traffic across 25% of the central Manchester area.

How did their AI technology help the UK’s Covid-19 response?

When lockdown began in March 2020, the company responded to a national government appeal for help in evaluating the impact of its policy across the country. By producing detailed traffic flow analysis, Vivacity Labs has been able to provide quantitative data to support Covid-19 response measures.

The data revealed that as the number of vehicles on the road dropped to levels not seen since the 1950s, there was a marked increase in cycling activity. The company’s ability to differentiate between vehicle types and activity over time, meant it was able to offer measurable data for the Department for Transport (DfT), to construct new transport strategies.

Impact on the business

Although its sensors hadn’t been designed to record person-to-person interaction, it became clear that this element of the business could be adapted and become incredibly valuable in the measurement and analysis of social distancing data. The underlying flexibility in Vivacity Labs’ video analytic technology meant that it was not only able to interrogate social distancing in action, but also compare this behaviour with previous data. The company was able to confirm that there was 95% less person-to-person interaction at less than two metres, in city centres. This tracking ability will help record further changes in behaviour over time, as lockdown continues to be relaxed.

Mark Nicholson, CEO of Vivacity Labs, says Innovate UK has now given the company a grant to research further and build more detail in time and distance interaction: “We were among the first to be talking to the DfT about compiling transport data and we were early respondents to measuring social distancing from various data feeds.…Three months ago we wouldn’t have dreamed of using our data in this way, but new-use examples emerge as time goes by. For us, this provides great validation, not just those set up for its intended purposes but also for the flexibility built into the system as well.”

Slingshot Simulations

Who are they?

Slingshot Simulations Ltd has developed a user-friendly, web-based service that harnesses the power of the cloud to create an ‘internet of simulation’ platform. Digital simulations or ‘digital twins’ of real-world systems enable organisations to test out scenarios before putting them into practice. In 2018, Dr David McKee, CEO and founder of Slingshot Simulations joined the Enterprise Fellowships programme, to accelerate his product development with the Hub’s funding and specialist mentoring. Since then they have thrived in the simulations sector and have become of the first UK companies to join the international digital twin consortium.

David says: “Our overriding mission is to democratise simulation. That’s about facilitating people, businesses, decision-makers to make rapid (in days not months) decisions that are SMART in an easy-to-use and accessible way using a no-coding approach. We focus primarily on the mobility sector showing how people and things move around, not only within city centres but within buildings.”

How did their AI technology help the UK’s Covid-19 response?

Using simulations for Covid-19 was a natural progression for the company. Digital representations of social distancing within and outside a building can be measured and simulated with their patented technology. Following recent funding from Innovate UK, Slingshot Simulations will be working with local authorities and businesses on the ‘Breathing Cities’ project. This will be analysing the relaxation of lockdown and the subsequent impact on the local community and pollution levels in Leeds.

The new project will focus on finding sustainable methods for increasing and maintaining resilience to the influx of traffic and people returning to. At the end of the Breathing Cities scheme, Slingshot Simulations will showcase a 3D animated digital twin that will inform future decisions on Leeds’ clean air strategy.

Impact on the business

The company has learnt a lot from lockdown. “Before that, we did not have any 3D visualisation or rendering in our product offering. Now it’s a core part of our business.” says David. The company has also learnt that it can operate virtually. All eight staff have been able to work online from home and contribute creatively throughout that time.

Finally, David notes: “We’ve learnt the importance of story-telling as part of our business. Not only between us and the stakeholder but between the stakeholder and their contacts… we’ve learnt over these last few months that being able to create dynamic scenarios is important and really works.”

Humanising Autonomy Ltd

Who are they?

Humanising Autonomy is a predictive AI company that helps improve how self-driving vehicles and systems interact with vulnerable road users.

The company’s team of behavioural psychologists and technologists look at cyclist, pedestrian and wheelchair user behaviour in order to predict what their next move will be. Its technology quantifies a range of data points that include a pedestrian’s intention to cross the road, if they are looking at their smartphone and if they have made eye contact with the driver. The technology provides a similar intuition to that of a human driver with software that plugs into existing navigation hardware and software, of fully autonomous and semi-autonomous vehicles.

In February 2020, Raunaq Bose, CTO of Humanising Autonomy joined the SME Leaders Programme to develop the skills needed to scale up the business.

How did their AI technology help the UK’s Covid-19 response?

Covid-19 has enabled a new use to be made of Humanising Autonomy’s behavioural video analytics software. The company will be carrying out Innovate UK funded work with Transport for Greater Manchester to better understand social distancing implications and assess how passenger behaviour is affected by the pandemic. By overlaying pre-existing CCTV infrastructure with the firm’s predictive AI software, urban planners and designers gain increased awareness of how their infrastructure is being used.

Raunaq leads the development of the company’s core technologies. He says: “We want to enhance the safety of staff and improve public transport interchanges. For example, to enable social distancing we are trying to find hotspots and bottlenecks in a station. You can create barriers to make more space or encourage alternative routes…This principle can then be carried out to junctions outside the station as well.”

Impact on the business

Humanising Autonomy did not have to pivot its business model with its Covid-19 work in Manchester. The modular nature of its system means that it can retrain AI parts rather than having to build a new system from scratch. Raunaq says: “You have to remember that people drive differently in Manchester than Paris or Tokyo. We have set up our technology so that it is flexible in various contexts...We believe in bespoke systems, our technology has to fit the city, not the other way around.”

SIGN UP FOR OUR NEWSLETTER

Subscribe to our mailing list and get updates!

Do you want to become a member, a mentor or just want to support us?

Find out more >