Safe, reliable, affordable and equitable transport is critically important for communities. Myra Blanco from VTTI tells Martin Cooper how using data is the key to enabling people and goods to move more effectively.
Virginia Tech Transportation Institute (VTTI) is one of the world’s largest driving safety research facilities. From its modest beginnings in 1988 with a dozen faculty staff, VTTI now employs over 400 people. It continues to be a defining force in using data to make vehicles and transport infrastructure safer, more efficient and kinder to the environment.
Myra Blanco is VTTI’s Chief Growth Officer (CGO). An engineer by trade and training, she’s passionate about data and using it to make transport safer, more sustainable and equitable.
‘Motor vehicle crashes continue to be the number one cause of preventable death for teenagers’, she says. ‘My son is a few days away from getting his learner’s permit. I’m interested in peace of mind; imagine how you can use data to train teenagers to manoeuvre the roads safely.’
Moving on, she says: in the US, we’re making a lot of progress, but we’re not providing training for all teens. Think about bias and equality in data — we need a system that’s for everybody. Not just the people who can afford it. The question is, how can we use all this beautiful data…how can we provide a system where all teens in their various communities receive training? How can we be fair?’
Different forms of poverty
This, in a way, seems to be a neat summary of Myra’s take on data and its potential for good. We should be able to use data to reduce different types of poverty — not just financial poverty. Safety poverty describes only the wealthiest in society having access to the safest systems, products and solutions.
She explains that in America, for example, food poverty is a real problem. It becomes uneconomic for businesses to serve the very poorest neighbourhoods, and as a result, people in those communities don’t have access to fresh and healthy food.
‘Transportation is the big equaliser’, she says. ‘We need a safe, reliable, affordable and environmentally friendly transportation system to move goods and people. And we need to help everybody — not just those who can afford it. And not just for people who live in big cities.’
Gathering data
Central to VTTI’s work is the pioneering idea of naturalistic driving. This approach involves fitting vehicles for research projects with sensors and cameras to gather accurate data from actual drivers as they traverse real roads.
‘When we talk about naturalistic studies, it’s about seeing what people do in real life. We’re filming people and collecting data when they do revenue-producing routes, during their daily commutes, or vacations’, she says.
‘We’ve collected over 70 million miles of driving data’, Myra explains. ‘That’s over seven petabytes of information. That covers passenger vehicles, tractor trailers, coaches…motorcycles. When you have that much data, you can start putting a system together because you have representation of all the actors on our roadways.’
Ensuring data represents the group being studied is a prime concern for VTTI. That’s because, Myra says, different groups will use and interact with their vehicles and roads differently.
‘We want to define research questions and how we’re going to answer them’, she explains. And the data needs to be representative, too. For example, we can’t just have early adopters. Early adopters might be different from others, so we start creating categories — age groups, different regions with varying laws of traffic, summer versus winter…we need to ensure what we’re studying is broad enough to make inferences — not just extrapolating.’
Diversity is the key to safety
Expanding her discussion of representative data, Myra says: ‘Imagine you were producing a technology and it had never experienced adverse weather. That would be a biased analysis and have safety implications if the technology were produced. What if a new technology is never tested with senior or teen drivers? They are completely different in behaviour, performance and reaction times. To avoid testing with bias, we need to understand the end user. And my personal view — I’m a human factors engineer — is that we should start designing the test with end users’ capabilities in mind. That helps us avoid bias in the data and the research design.’
Having diverse teams also helps VTTI’s project group spot biases hidden in research methodologies. As evidence for this, she explains that pregnant women tend to have very different and changing dimensions from women in general and certainly from men. Who better to spot this potential bias than a woman and, ideally, a woman who had or is due to have a baby?
For you
Be part of something bigger, join BCS, The Chartered Institute for IT.
‘We need to bring those lessons and experiences from our unique backgrounds’, she says.
Building these diverse teams, Myra says, is essential but challenging. ‘You have to be committed. Many people say they want a diverse team, but — the data doesn’t lie! So, ask: “Show me how diverse your team is. Show me how you are collecting data and how you’re using [the diversity data]. And, who is looking at the [diversity] data?”’
This begs the question: why go to the trouble of equipping vehicles with sensors and cameras? Why not just use simulators and digital twins to gather performance data?
‘You can do a lot with simulations’, Myra says. ‘But what do you base your simulation on? You need to base your simulation on real life. You can do a lot with naturalistic data to create more realistic simulations. And maybe you move from a simulation to a naturalistic environment to further test a hypothesis.’
From the dock to the door
Myra is also keenly focused on how transportation networks are used. She’s the CEO of Dock to Door (D2D), a coalition of over 85 partner companies focused on improving supply chains and models. The project aims to improve equity, security and prosperity in remote and rural areas. Ultimately, the group seeks to ensure everyone has equal access to goods.
‘We envision a system where goods can move seamlessly from a dock to a person or a business’ door and back again. It would be fully connected in terms of information and also resilient. We don’t want a supply chain that only works when everything is perfect.’
‘Right now, the supply chain has many links, and those links aren’t interconnected. We’re looking at a federated data platform that can connect all those links, from a container in a port to a package at a door. If you have that data, you can make decisions — you can create corridors and alternatives that will, for example, reduce carbon emissions.’
Such a data rich transport network also promises profound possibilities for organisations moving goods about. For example, moving vaccines might need the fastest route possible; moving non-perishable goods might need the cheapest fuel.
‘Right now, we’re doing this blind’, Myra says. One organisation might have data on their piece, but they are hugging it and not sharing it. As a third party, we want to protect that data...to present the minimum data set for the maximum good.’
Summing up, Myra says: ‘Data is power.’ And as we head towards a more AI driven world — and a world where AI might be doing the driving — this quote will most probably age very well.
