Utilizing Real-time Insights from Connected Vehicle Data

Utilizing Real-time Insights from Connected Vehicle Data

Carl Novelli, the Assistant Vice President of the Public Sector at Wejo joins The Infrastructors and shares his journey into the industry, and talks about his passion for the technical aspects of his job. Carl explains what ⁠connected vehicle data⁠ is and how Wejo harnesses this information to make roads safer.

Wejo⁠ is a data, analytics, and software-as-a-service provider that analyzes connected and electric vehicle data to create real-time insights. With this information, they give businesses and organizations across a variety of industries the power to innovate, drive growth, transform communities, save lives, and improve the world we live in. This is all made possible by data from more than 13 million active vehicles.

Episode Transcript

Scott Stanford:

Hi, everybody, and welcome to The Infrastructors, the premier podcast for engaging conversations with influential thought leaders in AI, tech, government policy, and smart city innovation. Today's guest is the Assistant Vice President of the Public Sector at Wejo. We're talking to Carl Novelli.

Carl, thank you so much for joining us, my friend, all the way from London, huh?

Carl Novelli:

Indeed, and thank you very much for having us. I'd love to say it was nice and sunny in London, but it's raining. It's cold.

Scott Stanford:

I love it, man. I love it. Can't get to London enough. The food is fantastic. A little pricey for my taste. But you are the Assistant Vice President, Public Sector at a company called Wejo.

Carl Novelli:

Yes.

Scott Stanford:

You guys deal with connected car data, smart mobility. Carl, first, as a guy growing up, when did you become interested in this industry? Is it high school? Is it through college? How do you get interested in this type of business?

Carl Novelli:

I think that's a great question. I'd love to say that I came straight out of high school, and this technology was already here. I would be lying. My background is absolutely in technology, in servers and IT and desktops and all those kinds of things. And then I personally moved into automotive and telematics and that kind of space. And I've got a passion for cars. I love trucks. I'm right now in the process of spending too much time looking at Tacomas, which I want to buy ASAP.

Essentially, it's tying all of that together, and that's where it gets really exciting. I love tech, I love the technical aspects to the job, but I love working in the automotive industry, but also the really cool thing about this job is I get to work with transportation individuals, the folks that are out there fixing the roads. I grew up playing with Lego, those building bricks and building it, and I clearly wish I'd have gone into transportation, went into tech. So now I get the best of both worlds.

Scott Stanford:

I love it. And the name is interesting. Where did you guys come up with the name Wejo?

Carl Novelli:

Absolutely. In fact, we either get asked, "Is it Weho or Wee-jo?" So Scott, you did really well with Wejo, tick to the box there. It's actually short for We Journey. We lost that headline, but that's essentially where the name came from, which, hey, it's great. I think we should keep that line.

Scott Stanford:

You could have went with We Journ. It would have had the same effect I think, no?

Carl Novelli:

Absolutely.

Scott Stanford:

Where did the idea for the technology for this company begin and just explain to me what you guys do. What is connected vehicle data exactly?

Carl Novelli:

A great question in its own right. Depending who you speak to, people may have their own interpretation of that. When we say connected vehicle data within Wejo, we mean vehicles providing information via telemetry out the vehicle and then we are harnessing that information and providing that information out the door to, for example, departments of transport traffic engineers to make roads safer. Fundamentally that's what it's doing. So imagine your vehicle that you drive, all of those systems in that vehicle, all of those sensors are collating information on the road, on the roughness of the road. If the vehicle is bouncing down the road, if you are slamming your brakes on, if you're driving really quickly, all of those sensors are capturing this information. Imagine you are a traffic engineer and you are looking after a city.

And that city, you've got cameras in locations, you've got capture devices here and there that can see sort of snapshots of what's going on, but now you've got the vehicles themselves so you can have that bird's eye view looking down on the city and seeing where are folks slamming the brakes on? Where is their safety issue? So that's what we do. We essentially work with vehicle manufacturers or OBMs and we help them harness the data and we provide that data out the door to end users like engineering firms, those kinds of folks.

Scott Stanford:

Because you say data, I'm going to go with data instead of data for the rest of this conversation, Carl, because that's the kind of guy I am.

Carl Novelli:

I appreciate.

Scott Stanford:

Do you have a number on how many active vehicles that you guys have data from at this point? I assume it's going to be in the hundreds of thousands, no?

Carl Novelli:

Yes, absolutely. If we were to look just at the United States alone, right now we've got say about 13.7 million vehicles.

Scott Stanford:

Wow.

Carl Novelli:

We collect information every one to three seconds. A huge number of vehicles. You could quite comfortably say we see between three and 7% of all journeys on US roads.

Scott Stanford:

Wow. That is insane. And there are people sitting there and the computer is just wrapping up all this data and sending it right back out?

Carl Novelli:

Yes, to a point, I think what's really important is this data is fully consented and it's anonymized data as well. So we are not looking at this and saying for the traffic engineering and these safety kind of use cases, how is Scott driving his specific car? It's more the mass of the vehicles. Is there a pattern? Is there a trend? Is there lots of folks on an on-ramp hammering the accelerator because they have to, otherwise they're going to have a problem. So that's why there's a piece there that ties it together.

Scott Stanford:

Why is connected car data different from traditional mobility sources?

Carl Novelli:

Again, really good question. If you think back to the journey of traffic data and that type of thing from your guy or girl by the side of the road with a clipboard doing a classic traffic count, you would send them out from like 8:00 AM till 10:00 AM, four till six, and they'd come back and say, at this intersection I counted X number of vehicles that went that way, X number of vehicles that went this way. And that's where it all started. We talk to cities now that are using that data to make decisions on planning, safety, all these type of things.

I had a conversation with a gentleman up in Colorado and I said, "Oh you use that?" He said, "Yes, this is how we capture our data." I said, "Great." He said, "Yes, but we only do it when it's sunny." I said, "Okay. That's interesting. But you're in Colorado, it snows a lot."

So if you think, going back to our example of you've got all these vehicles driving around the road 24 hours a day, 365 days a year providing that information, you've now got really good solid basis to form decisions on. So that's one example of where the data captures come from to where it is today. And there's other data sources. There's things like mobile phones get used quite a lot to capture this type information. And again, these data sources are great because the more that you can base your decisions on information, the better your decision's going to be. We are a layer to the story fundamentally.

Scott Stanford:

So, all this data gets taken in, and correct me if I'm wrong, but it's all to make the roads safer, reduce congestion, help driving become more cost-effective, maybe help the air be cleaner. Do you have to be a data expert to digest all this information?

Carl Novelli:

The key here is the folks that aren't data experts. It is the people that have the questions or the people that have the problem. If you picked out one of those bits around making the roads safer. As you know the infrastructure bill has released a huge amount of money that can be used for this, which is fantastic, but where do you spend the money? Which location do you need to improve? And that's the question. So the question would be, we've got the money, we know unfortunately that deaths on the roads are going up, people are driving faster, people aren't using their seat belts. Driving is getting worse and we can speculate as to why and all those things.

But if you can then use the data that we have to identify hotspots where there are issues. So let's say an intersection where loads of folks are slamming their brakes on, which you can see in our data all of a sudden that that's something you don't need to be a data expert to see a pattern. If I make a simple change, it's signals to re-timing or something along those lines, then the following day I look back in the data and there are folks still saying, are my brakes on, yes or no? And that's how cool this is because it's captured all the time, you can make instant decisions.

Scott Stanford:

You talk about the slamming the brakes and the US here, it's such a crossroads where there's an epidemic of fatalities right now on the roadway, which coincides with this massive investment and a strong need to totally transform our transportation systems. Where do you see Wejo fitting into the story? How will it partner with others like a Rekor to get there?

Carl Novelli:

This is the key here, and I use this analogy a lot. We're not the unicorn, we're not here to say we've got all the answers. We can fix everything because whoever tells you that is lying. We are part of the story. As we've mentioned, we've got all these vehicles out there providing this level of information. The great thing about Rekor and the platform is it essentially allows the information to be viewed in a really clear and precise way. And I'm not a trained traffic engineer, but I can use the Rekor platform and start to see trends. I can see patterns, I can see there is something going on here, we need to focus attention and I don't need to be a data scientist to do that. It's the bridging of the gap between this huge volume of data which is core and has massive potential. And answering a question as well.

Scott Stanford:

The electric vehicles obviously are just becoming more prevalent by the day. Every automobile company now has their own brand of electric vehicle. Can you explain how Wejo shows areas of our cities and suburbs where there might be the greatest EV demand? How do you guys do that?

Carl Novelli:

Absolutely. And I think this is exciting because one of the key things, if you look at the infrastructure act, is making sure no one's left behind. So the equality and really driving that home. So if you look around charging infrastructure, if you are a city, an NPO, a DOT, whatever it may be, well the first question is there's some charging infrastructure in and around, but how do we improve that landscape? So you pick up a newspaper and I think it's getting less and less, but people are concerned around that big old term of range anxiety. I'm in my electric vehicle, can I get to a national park? Can I go over landing, hang out for a couple of days in a cool place and get away from work and all those cool things I could in a traditional combustion engine? A piece of the missing puzzle here is, but where do we put the charging infrastructure to alleviate that?

So how do I do that? Well, if you start back from the bases, you want to get that understanding of where folks are generally going? What do those journeys look like? And I use the national park example because this is something that they're really focused on right now is opening it up to more EV users. Well if the national parks in the middle of nowhere that normally they are because they're beautiful places. If there's charges there but the closest city is a day away, there needs to be infrastructure along the way so you don't have these deserts in between.

So by understanding where journeys are starting and ending, you can get that kind of, where do we potentially have a desert? If you see where EVs are going today, that starts to tell the story as well. If they're not going from the city to the national park, do we have a problem here? You can see combustion vehicles are doing that big journey and that's great, but if there's no EVs going there, there's something going wrong. So now you can see this in the data. If you can see the EVs are going so far, well maybe we need to put charging infrastructure there. So now fast-forward, you've looked at it, you've put charging infrastructure there, amazing. You now look at the data again. Now they're going so much further. That's how to use the data to make these informed decisions.

Scott Stanford:

So it's just wild the way you guys look at things so differently than the rest of us. You talked about Colorado and only getting that data when it's sunny out, the insights that are generated using that CV data help emergency services and DOTs as well, helps them better understand that human behavior during those weather events. So how did Wejo data help? I know you guys, with the Indiana DOT, helped them make informed decisions on signal timing and hurricane evacuations.

Carl Novelli:

Absolutely. I think not so much that the hurricanes up in Indiana I hope, but around the signal timing piece. So again, it's understanding where vehicles are coming from, where they're going to. So some of the work that the folks up at Purdue University have done, Darcy and the team is incredible. Using our data and essentially putting it into dashboards that then can be used to see, do we have a problem? Is there a downstream blockage? Is there something going wrong at a light? Scott, you probably experienced yourself, you're driving through a city and you then suddenly think, hang on, why am I getting all the reds?

Scott Stanford:

It's not my lucky day.

Carl Novelli:

That's happened to me, and then I question everything. I think I'm not buying a lotto ticket today. So by using the data to see what the traffic flow looks like through these signalized intersections, you can then start to see, well having the person stop once, that's fine, that's okay. If they're stopping repeatedly, maybe there's an issue. If at one intersection you see at your rush hour, which has faded, Covid's happened, is rush hour now 10:00 AM when everyone's going to get Starbucks? I don't know.

So actually is the signal timing set up more for traditional rush hours and now it doesn't quite fit? So they're using the data to identify where they have issues to then make change and then do change and see has the change been effective. That's this kind of signal timing. You mentioned hurricanes, obviously any of these big weather events are awful because fundamentally they risk life, they take life, all those kinds of things. You are familiar with, obviously, the snowstorm, there was a big stay at home, don't travel. So you could see essentially the impact of a storm rolling across the US in our data. What you were seeing is fundamentally vehicles at travel times increasing where vehicles are now, the average speed on a freeway is 70. As an example, storm rolls through, everyone's going really, really slow and you can see that in the data in that live feed.

So it's having that at your fingertips to make an impact. Have you been down to Florida, Scott? I'm sure you have.

Scott Stanford:

Sure.

Carl Novelli:

You see signs posted, the hurricane evacuation routes. Going back to that analogy of the guy or a girl who stood by the side of the road with the clipboard, seeing where folks are going. If there's a hurricane, you don't want them doing that, you want them out of the way. So what you can do with the data is you can really easily see was that evacuation route used? Was it used to the best? It is a case that folks are not using that evacuation route. They are going another route. So do you then need to adjust mapping and rooting and things along those lines to make change, to get people out the way quicker because that's the key here is making sure folks... And it goes back to the point originally, it's using the data to make informed decisions to fundamentally save lives.

Scott Stanford:

Carl, you guys work with retailers as well, right? How do you help retailers gain a better understanding of customers needs and behaviors and enable them to tailor strategies and services all through smart mobility? How does that work with retailers?

Carl Novelli:

I think the retailers are exciting and interesting, and it comes back to that basic understanding where folks are going, where they're driving and how they're driving. So that drive behavior piece. And again, understanding not just one person but the masses because that's what we are looking at here. So you are exceedingly successful already, but imagine now you are diversifying your portfolio and you are looking to put some money into... What's your favorite food, Scott?

Scott Stanford:

My favorite food is pizza, my friend. Got to go pizza.

Carl Novelli:

With the surname Novelli, I'm definitely in agreement. So imagine your diverse manual portfolio. You want to set up a pizzeria and you know that the pizzeria that you want to set up is going to be the best thing ever. And you've already chosen your city and your city is Denver as we were talking about Colorado earlier. What you probably want to do is identify those busy roads and identify, if I'm a retailer, you could look at other retailers in that similar space and say where will people come from to go to their locations? So that origin to destination. If you looked at your top five pizza places, where are folks coming from?  So, there's a sweet spot here where there isn't a pizzeria right now, but there're lots of people traveling from there to another pizzeria.

So now you're starting to use it for locations. This is potentially a good spot to put in my pizzeria. Now I imagine the other example of you're starting to build your pizzeria, you've got some money for advertising and you want to go put up a billboard. You want to understand, well where are folks queuing? Where are they stuck in traffic? Where do they slow down? Where do I want to put a billboard up? Is it the traditional, oh we know there's always a traffic germ there or is it actually on the back of how patterns have changed and things like that? Is there actually down the road, another 500 yards, where the land has a lot of value and a billboard go really well there? So by understanding where people are going, how that driving behavior allows you to, again, it's all about making these informed decisions on a really solid basis of information.

Scott Stanford:

One that's near and dear to my heart, insurance premiums. Does your data, can it help reduce insurance premiums? Can it help insurance companies? Can it help their customers? Everybody, benefit from the data you guys compile?

Carl Novelli:

Yes, absolutely. So this is something that we're starting to work on and we're starting to partner with some OEMs on this. Again, I think the important thing here is around people's privacy because that is very near and dear to our hearts here at Wejo with... One of our tag lines is data for good. So you imagine, say for example, your vehicle and you give permission for your information to be provided to an insurer?

Absolutely. If you are happy for that to happen, that is something that we are working on. And then it goes back to, we have it in the UK, the black box in your vehicle that would say, well I'm definitely a safe driver, I'm really good, I never speed and actually I only drive between these hours. I don't go out late at night. All these kinds of things. As long as you've given the permission, then absolutely this is something that that data could be used for. But it all comes down to the consent model. We are not here to use this data in the wrong way because it's such a precious asset and it comes back to you, the driver, what do they want to do with the information.

Scott Stanford:

It's almost like when the insurance company asks, well we're putting together your premium, how many miles do you usually drive away from your home? And if it's an older person, it's not that much, well I do 20 miles a day from my house. But I guess your data actually gives it to them in real time, factual information.

Carl Novelli:

I use my example quite often. I've got two cars out on the drive and I don't drive anywhere. In all honesty, we get our food delivered on a weekly basis. I drive my car to the airport more than it goes anywhere else. My wife's is used all the time. So does it make sense that the insurance that we pay on my wife's car and my car is the same? You always want to touch wood when you say this, but the risk of me getting into a wreck is probably lower because I'm simply not on the road now.

Scott Stanford:

So Carl, how does Wejo support future autonomous vehicle development, the testing and operation, how do you guys get involved with the support of those programs?

Carl Novelli:

This is a really exciting space. You mentioned that you like pizza early on, have you seen any, you probably have, the autonomous delivery of a pizza.?

Scott Stanford:

Yes.

Carl Novelli:

It's that kind of stuff. And I think the key here where Wejo sits in is to improve autonomous driving you need to know what's going on the roadways. You need to know what's happening. So how do you do that? How do you feed into this? How do you allow the machine, the AI, that's driving these vehicles? You need to train it, you need to give it examples of this is what a freeway, this is how it is being driven. So by using the data that we capture 24 hours a day, 365 days a year, actually it gives a good baseline and it feeds into the modeling aspect. But now imagine you're a DOT that is looking at autonomous driving on your roadway.

Actually, you want to identify where those roadways have an issue for, not just autonomous driving but General Joe driving down the road. If the infrastructure's not great and you can see that in the data that helps to support. So fundamentally that's how the Wejo data is being used as that baseline to help train the models, to help understand the roadways and help see the impact of events. So we kind of have spoken around some big weather events and things like that and what does that mean for roadways. But imagine the unfortunate situation where there is an accident on a roadway. Then how do vehicles drive? What happens? What is that ripple effect? So that's the information you can see in our data.

Scott Stanford:

And then listen here where I live in the New York area, you see that effect every day where if somebody breaks down or there's an accident, it's a ripple effect for miles and different roads and different on-ramps. It's unbelievable what you guys can pull in from the data you collect. And Carl, before we go, we always have these conversations talking about the autonomous vehicles mainly for like you said, deliveries of things. We're seeing these different vehicles slowly creep in. We were talking with another group of friends about how long will it be available and how long before you yourself trust getting into the back of a car without a driver in it? How far down the road are that happening where our trust in the system and the technology where you're going to pop into the back seat, take me to the airport and you're fine with there not being a driver in that car?

Carl Novelli:

Listen, in some cities it's already there. It's been rolled out. I've not experienced it, I've not had that and I'm waiting for the moment of either hailing a cab or ordering a cab via one of the apps and it turning up and opening the door and being, "Oh okay, right." This is that crunch time. The interesting thing is if you look at one of the problems on roadways and I do it, we will do it, distracted driving, that type of thing, if you take that out, autonomous vehicle is not going to get tired, it's not going to answer its phone.

It's not going to see the pretty girl as you drive past. It's not going to see the dog as you drive past and all a sudden look over there. So I think if you look at it from that logical standpoint, the future does make a lot of sense. But then I'm also of the, and I mentioned a tricoma earlier on in this conversation, I'm a big fan of over landing and going out middle nowhere, taking the truck, camping, all that kind of stuff. That's cool to me. And I love vehicles and all that kind of stuff. So I think there's going to be a place for both. You mentioned New York, if you imagined Central New York and it's the classic thing, unless you live in New York, oh my goodness, New York cab drivers are terrifying and I know they're not. I know they're the safest drivers ever. I don't want to get hated next time I go to New York. But imagine that those vehicles are now autonomous and all driving at the same speed and it's very, very safe.

Imagine what that now looks like from a safety standpoint. No one's getting tired, no one's getting annoyed, no one's getting hungry, that hangry comment, those kind of things. That's now not a fact. Will I have that moment when that autonomous cab turns up when I open the door and think?-

Scott Stanford:

I'll get the next one.

Carl Novelli:

Yes. Well I think it'll be really, really exciting. And I'm in this world of transportation, I see absolutely the benefits and the good. But I know that's a real wishy-washy answer so I apologize, but that's kind of where I am on that moment of opening the cab door.

Scott Stanford:

Funny. Carl Novelli from Wejo. Carl, listen, thank you so much for spending a little time. Listen for your appearance on the show today. One Toyota Tacoma will be delivered to your home within the next couple of weeks. So we got you covered, my friend.

Carl Novelli:

I appreciate that.

Scott Stanford:

And with no driver, so you're good to go. Carl, thank you my friend, fantastic stuff and just such interesting information and keep doing what you guys are doing and I know in the next five years we're going to be so far ahead of where we are even right now... Actually, I think about all these futuristic movies with the robots and the flying cars and everything else. I think in 2023 we probably all thought we would see it by now, but I think we might be a little ways down the road. But it's coming, man. It's definitely coming. I know that.

Carl Novelli:

It's coming. Listen, the Jetsons, that time has passed which I think we were all hoping for those hover cars, but as long as we make the roads safer, that's got to be the number one goal with all their stuff-

Scott Stanford:

Jetson.

Carl Novelli:

We've got to do it better.

Scott Stanford:

Carl, thank you, my friend. Thank you very much and continued success, bud.

Carl Novelli:

Thank you sir. Have a great day.

Scott Stanford:

Take care my friend.