In this video interview, Nira Dynamics Product Strategist Björn Zachrisson and our Managing Director George Cambanis discuss vehicle data products and their usage for multiple use-cases.
Hello. This is George Cambanis, one of the co-founders of Mobito. And we're here today for another data provider showcase. Very happy to have with us Bjorn from Sweden. And Bjorn is going to talk to us about Nira and their very interesting connected vehicle data and derivative data products. So, Bjorn, thanks a lot for your time and for connecting today.
If you want to start with an intro on NIRA and perhaps where you guys coming from background of the company and what's your key focus? That would be great.
Yeah. So my name is Björn Zachrisson. I'm working at NIRA Dynamics and has been doing so for the last seven years. And Nira, I mean, we are a tier one supplier, so we're almost in the motor sector. But we are not a car manufacturer, we're not in OEM, but we are tier one. So we've been supplying software for the vehicle industry for the last 20 years or so.
So I think it's about 75 million vehicles that has our software in it. But those are the non connected vehicles. So these are the in vehicle services for the vehicles. But that's also where we have learned a lot about the the vehicle behavior and the tires since our core product that has been and still is the tire pressure monitoring system.
And we do that without sensors. So we do the sense of fusion, as we call it, meaning that we create new values from existing sensors and we do processing inside the vehicle and that is what has been the way to get us where we are now, where we utilize our knowledge in order to create information about the road that we then extract from the vehicles and put it up to the cloud to create these aggregated views of road information in terms of general road safety, asset management, but also the slipperiness for mainly winter maintenance.
But we're also doing all the investigation how well we can do it during summertime and I would say, I mean, our core business is still in the vehicle industry, but I think a lot of our focus now is on the road information gathering. That's where we put a lot of effort in because that's our new research and development areas.
Got it. No, that's very interesting and very useful to know how the company started this business and what's the background. And yeah, perhaps it's worth giving an application to the connected vehicle data market, which, as you just mentioned, is this area where you're now laying more focus on. From our point of view, from the marketplace, a mobility data marketplace point of view, we see the the connected vehicle data being one of the more exciting and impactful segments of the market.
And of course, this has to do both with the new data availability into the market as we have more and more connected vehicles, but also by the big demand that many use cases that are already using and waiting to use such data. Perhaps worth clarifying as you started doing that, there are different ways of accessing vehicle data. One is of course from the OEMs directly and any sort of such an ingrained connectivity of the vehicles.
And the other is by companies that have the ability through some sort of after sale devices, dongles to do in a sense, install connectivity into a vehicle and start capturing signals from there. That will be the one key distinction. The other key distinction from what we're seeing is that there is data that pertains to individual drivers and individual cars, which is not anonymized.
And this can feed use cases such as dynamic pricing for insurance pay as you go type of models. But at the same time, we see very, very interesting development in the anonymized data from vehicles that going to be used for traffic analytics or perhaps road quality, like you mentioned. So with that background, can you help us understand your positioning in the connected vehicle data market?
Yeah. I mean, I think we started out back in 2013 when we did the exploratory investigation using the aftermarket devices, knowing that only then we would be able to put our software inside the Volkswagen Group vehicle. So once they noticed he had scored a Porsche. And in the beginning, I mean, we had a lot of struggles just getting into the function properly with the hardware and everything that you need to do that.
But I think we got it pretty decently working. But it's always a challenge when you have these aftermarket devices with with a maintenance and having them plugged in and not being removed and all the challenges with the people driving. So I think that was an excellent way of getting started. But we knew from the beginning, I mean, it's not a scalable solution and it's very hard and expensive to maintain a proper fleet that if you want to have massive coverage, which is the whole idea behind the things that we are doing.
So in the year 2019, we we have a corporation with a Dutch company. It kind of went from having a thousand vehicles, up to 50,000 vehicles with aftermarket devices because they already had an aftermarket device that was installed from there from the dealership in the Netherlands. So we could gather data through them. And that was the start for then the real launch back in 2020, the end of the second half of 2020, when we got the vehicles on board on the factory vehicles.
So when the user buys it and accepts to send the data or decides not to send data, we don't get it, of course. But once you accept Sunday that we get it from all the vehicles that we request and we kind of try to keep a good balance in the amount of vehicles we get to to optimize the service and also the cost that it comes with having a large fleet.
And from the beginning, we have been very thorough in keeping the air personal integrity intact. So not jeopardizing any kind of personal data. So from from my point of view, GDPR is a relaxation of the requirements that we have on on us to do that, make sure that the car's not traceable. I mean, we go to great lengths to anonymous data with the purpose of being able to extract data from as many vehicles as possible and put it up as a more or less complete map layer, especially for the asset management use case.
The window maintenance is harder because a complete map means that you need pretty much three vehicles per inhabitant in the country. So that's kind of difficult to achieve.
Got it. I'd love to to look into the data, perhaps. And it would be very useful to have a quick demo of the dashboard before we do that. Since we are discussing as part of the mobile platform network, I would love to hear a couple of words from you on. How have you benefited from our interaction and from being part of this network so far?
Yeah, I think I mean, this area is still fairly new. I wouldn't say immature because it's been around for a couple of years. So I think there's a very large, large area with a lot of actors and a lot of data providers and a lot of platforms. And I mean, that's usually what you see, I mean, in a new kind of area.
And then there will be, I mean, fewer and fewer as time progresses that both provides data but also broker data. I guess it doesn't make sense to have 25 different companies in the end. I mean, it would be acquisitions and stuff and you would just have a will have. But working with Move It I think has been been great so far.
And I mean, you guys are an active company and you really push forward to connect them the data and also the buyers and understanding what kind of data that is that that's being produced. And I think that's the key in order to succeed that you also understand what it is that you are selling is having a platform to sell whatever can be kind of challenging.
So I think that's one of the strength with it. I mean really understanding the business and the data and the use cases and we've already gotten a couple of leads from I mean, from the corporation together with Moby-Dick that we are exploring together to further expand our corporation.
Super. Thanks, Bjorn. That's that's indeed very, very aligned with our perspective of the market. And I think we have been forced to have this more active approach and, you know, helping both sides of the marketplace, giving a lot of the nascent in these markets and the new use cases and the need for people to understand and to compare data availability.
But we're super excited to be working with you and to be very, very, very interested in the type of debt that you make available. And with that, it would be amazing if you can share with us anything you want to share. I know you have a couple of very interesting dashboards showcasing part of the data, so I'll let you decide which parts you could make.
You can showcase really briefly for the benefit of the audience.
So this is from the intimate inside. And this is from last December in 2021. In payroll outside of Stockholm or in Stockholm, depending who. Yes. And what you see here is the actual friction, not the forecasted or not the assumed friction. It's the actual friction of the roads which is a huge difference since we I mean, it can differ extremely in just small sections of the road and relying on single spot measurements doesn't really give the truth.
And this was not a day with precipitation or the day before. So it wasn't supposed to be slippery at this day, but as you can see here, you have the red meaning that it's slippery and then you have the yellow orange. That is kind of I mean, in between it's not good, but it's not super slippery. And then you have the green, which is good.
And we also see the friction distribution up here that's make a bit more simple, have what we call a good friction, and that should not be confused with friction approximations. But if you look on the real friction for vehicles point seven and up, that is what a good friction is. If you're looking from the OEM perspective and that's when you are allowed to do freeze, the more or less on the outbound, for instance.
And we see that there's very few spots where the good friction, but at this hour at 6 a.m. there's not that many vehicles out. But if you go for this here, 730, we see a lot more data and we see the distribution. It's weighted in the lower end. But this example is interesting because if we look on the highest road gas, meaning in this case we have eyeglasses, means pretty much the middle class one is high speed, high traffic density.
And this is one of the main arteries leading into Stockholm and out, of course. And these are all more or less never be slippery. And it's it is pretty good. I mean, you can see the distribution here. It's completely different in comparison to all the roads. And and I think this is interesting. And then if we switch or go in the different times, we can see that it becomes red on some places and we can just look on the slippery areas.
It might be easier to see and we can also see the distribution of of slippery as would be the amount of sleepiness in the different hours we can see the morning traffic was much worse than the later on during the day, but it was continuously press here. But if we remove this again, we can see that the entire area was more or less super icy, so it was hazardous throughout Stockholm, which it wasn't supposed to be because there were no precipitation and it was minus degrees or cloud all over.
This is one of those odd cases where it is leprous, but it wasn't expected. I think that's when you need a friction to get the information that you need to make. Do some actions in order to maintain accessibility and road safety.
Super. And you're maybe here for the sake of clarity in order for you to be able to get these actual measurements on a daily basis and even hourly basis and even more granular basis, you don't need to actually send any of your own vehicles, but this is made available to you by the underlying fleet that you have access to.
Can you please, you know, clarify that in order to also understand how can we activate and deactivate some new geographic areas we're interested in?
Yeah, I forgot about that part. So, I mean, we have about I think it's two, 3 million vehicles that are available for us to activate in the regions where we are active. And that means the not entire year but most parts of Europe. So Finland, the Baltic states, Poland and then going south down to the coast of Italy I think, and also Greece.
That's where we have the data active pending the U.S. and soon to be Canada. And then we can request more vehicles per area if we need more vehicles, which would be driven by the winter maintenance, we can pretty much just ask to to activate more vehicles or activate the new market. And then we would turn on the the data stream coming from those vehicles.
But we try to keep it again. I mean, try to keep it as efficient as possible to not overuse vehicles due to the I mean, the the excessive cost of the processing for us. So we try to just use what we need.
Right. And I think you already provided a number early on in our talk, but can you share with us I don't know. If you look at the whole of Europe, the potential available underlying vehicles that you could activate? And is there a range you can share with us how many such vehicles there are with this potential connectivity?
I think it's about 2 million on an annual basis of connected vehicles. And we try to use the latest software that we have out because I mean, it's being updated with versions of the in-vehicle software. So I think we now we are around two and a half million vehicles that can be activated in Europe to two and a half million vehicles.
And I think we are using 1.2, 1.3 million vehicles as a current.
Got it. Very clear. Very clear. And of course, this number goes up pretty much day by day as you have more connected vehicles hitting the market from your underlying OEM partners.
Yeah, exactly. And some of the categories we have already saturated needs where we don't think that we need more vehicles and some markets are still in the roll out phase, but most of them, I think, have enough to to provide the winter services. But there's also this challenge. I mean, if you have 10,000 vehicles and you have to increase it to 20,000, it's not like you're going to get double the coverage.
You might get a few percentages, extra coverage depending on how I mean, how much 10,000 is in relation to the road network, but I think the best example is in the UK where 2% of the road network carries 35% of the traffic, which means I mean the highway network would always be covered and if we had another 1000 vehicles, three or 50 of those would be on the network that we already have covered.
And maybe one follow on there. Let's say, you know, someone who wants to activate a new area, a new country, a new city. And what does that entail for you? Like how how quickly can someone get access to data from this new area?
So it's a brand new country and it's in with a region where we are active. So exactly where we have access to vehicles from from day one, it's about from like week up to four weeks depending on how how the workload is with within our team, but also our data suppliers workload. So it should be fairly, fairly quick.
Yes. Okay, super. So within a month time, someone could have our first mapping of the roads in terms of road quality. And from these scoring that you guys provide.
And maybe could also show something from the asset management side. I did prepare another demo, but this one is a a little bit maybe more complex, but we can try it because this is the roughness or the road state, the international roughness index and the choosing an area in Florida because I did a demo for Florida some time ago.
So I've chosen this road stretch in the rural region because it is kind of interesting because it's a this this colors in blue and green means that it's pretty good and red means that it's pretty bad. So this is the actual state of the roads, which is, I mean, something we can create instantly, more or less. I mean, as soon as you have a car driving over there or maybe two cars, we can have it.
The cover and try to ignore all the data on the side. This is the scientific part of the tool, but we can look on this one, which is the trend, which I think is one of the key features of the hazard management side, that being able to collect the data on a daily basis, we do daily snapshots because the road doesn't change that that fast in the duration rate, we can establish trends over time.
And this again, the cornering means that green it's slightly the negative and red means it's highly de energy and dark blue means that it's become better. So looking at this just this area, we can see that here, it's getting bad in the past rate and here it's getting good in the fast rate. And this one is connected to the more the deeper level of information.
But if we click on this one, we would get to the Google Maps. And I think this was the area where it had become better. I did look on these areas yesterday, so yeah, but the maps are from the summer. But we can see here that on this area you have a newly paved road on the part of this stretch which clearly indicates I mean that is become better and yeah it's a newly paved that should be good.
And I know this is a lot to take in the short period because normally this takes much longer time. But if we look on this area, we have a high the innovation was in this area. And if they come up here on the same road but a few hundred meters, maybe a few kilometers away and we see a completely different setup of the road.
I mean, this one does not look really nice. So that is one of the things that we we can do that we can early identify when traveling or the deviation starts and highlight those areas in order to be more efficient in your road maintenance. Find those areas before they really deteriorate and breaks down.
Super, very clear. So yeah, you can provide both of course, almost real time updates into the state of the roads. But then from this analytics, you can also see that the generation through time for the period of time that you're focusing in. Yeah. Yeah. And Bjorn, I think it's going to be super nice if you can hear a little bit more about the use cases of people and how they're using in different industries.
This data you already mentioned road maintenance, of course, which seems like an obvious candidate for benefiting from this data. But are there other such industries that you want to mention?
Yeah, we also work with the the OEMs of course. So we we warn the OEMs all this using they did a press release where they use our slipperiness in order to warn the drivers of sleepers up ahead. And I think it's also worth mentioning that when we talk about slipperiness, we don't talk about brake indication or high acceleration.
We talk about the actual friction while driving. And that means that you can get these warnings when it's I mean, some part is slippery, like a show on the map. Before with the friction, there was parts of the main road that was slippery. Those warnings you would get up in your heads up display. And I think that's a huge difference in in comparison to use abs or peaks only because pretty much what you get then is that if you have a high density of ABS enthusiasts, you would know that, okay, this is something going on in this area.
But if you get high density, I mean, we've already done those analytics by ourselves. If you get a high density, you don't need to look on the data. You can just look up in the sky and you pretty much see the snow pouring down. That's the only time when you get that real intense amount of ABS enthusiasts on on the highway because normally people don't break or accelerate hard on the highways, whereas we measure continuously during normal driving.
And that's what we can detect those small spots on the roads. And then I think the sorry.
No, no, no, just I think this is a great way to, you know, start speaking about their unique selling points and what differentiates and I guess data back to the use cases besides the OEMs and the road maintenance, do you see any other key applications of this type of data?
Yeah. I mean, insurance companies is one where we we have had some interesting we had some some talks that I mean, from from two angles. I mean, it could either be insuring them the infrastructure. So the infrastructure side being taken an insurance protecting them from claims from accidents due to their own state kind of and in order to to take the insurance or I mean, to insure an asset, you might want to know something about this asset, especially, I mean, within summer conditions.
And in in England it's clear that thing I mean, that if you don't if your roads are slippery during wetness, which is a challenge in the UK, you might not want to insure that kind of infrastructure or they might not want to pay them due to that. The roads are in in bad conditions. I mean, it's it is hazardous on that road and you don't want to be the guy insuring that road then.
So that's one of the use cases. And the other one would be if you had some do not going too much data. It was the way we have some corporation where we're looking to how drivers drive on different roads. If you have the there what you talked about the UBI kind of set up so the user based insurance, how you drive, you can also connect that to to the roadside.
How do you drive during winter conditions, for instance, do you drive safe? Would you drive challenging, which would also be something to weigh into the general insurance cost. So insurance companies are interested and I think that's one of the areas where we will continue to explore the opportunities.
Great. Thanks, Bjorn. It's it's really nice to hear about the different use cases and how they're using your data. And one follow up question here that could be interesting for the audience is the fact that, you know, essentially the core of your insight is a road quality score. And my question would be, given that this is a fairly regulated I don't know if that's the correct word, but there exists, you know, standard way and processes for different players and stakeholders to measure road quality.
How does your score compared to those other industry standards?
And yeah, I if I got a cent every time I got that question, I would be a rich man by now. And so I think there's I mean both of wind domain that as an act management, there's always these kind of certification in order to do proper measurements and that ways of doing things that you have to do a lot of things in order to be qualified to perform a certain index or something like that.
And I think that is changing. But it's I mean, it's not going to happen overnight. I mean, the nothing happens overnight. And this is a fairly conservative business. I mean, road has been around for, well, clearly 2000 years is rolling started building roads and they're still around the same rules, which is amazing, but they're not that great and I think great comfort, but I mean, we've been working a lot in Sweden and also in the Netherlands and Scotland and all of those countries have certain ways of measuring the road site, especially for the summer, and we wouldn't be able to create a class one higher classifier because that is when you have the laser quality more or less, which means that you can identify half a millimeter crack from a point 25 millimeter crack. But I also was on the PR conference two weeks ago, I think in in Italy, where one of the keynote speakers talked about that you can from these fast new equipments, you can get 4000 measurement points per square meter. And then he also said afterwards, but I don't know why and I think that's one of the questions that has been forgotten a bit, that why would we need even more high resolution data?
How would that make our decisions better? Is it very expensive to use those equipments? You have to have that driver. You have to have the hardware and you have to do it fairly rapidly. Well, with the high frequency, because if you rely on two year old high quality data, that's from one of the dots that we were working with.
They had two year old data were old saying that road was in great condition. I mean, it was pitch perfect. And then he was driving down the road and feeling like this is a perfect road. Then I think we are doing something wrong. So he looked in arguing and we said, or in our tools and our data said that this road was in poor condition.
He was getting poles and it was generally rough, whereas his two year old high quality data said the road was perfect. So I think there's I mean, there's going to be a change in the way of working as the cars. I mean, we won't replace that. The measurements have been before today, but I think that the industry will change in order to use more of our data, or at least to single out where they need to do further investigations and not so.
That was a long answer on your certification question, but we won't certify ourself as the current equipments do today. We won't have that accuracy, not with the current technology. Maybe if we combine with camera later, we could get close. But I don't. That's not really our ambition. Our ambition is to create the tools that really makes the life easier for the users.
So singling out the areas where where things are happening, where you want to do things in the maintenance actions, for instance. So I don't think it's I think it's a challenge to break the barriers in coming into the usage of our data. But from the properties that we've been running with the Netherlands, Sweden and Scotland especially, I think we have proven that our data adds value and really shows that using the vehicle data, connecting data makes it or enables faster, maybe better decisions.
I mean, if you have a measurement from the same day I would go for I mean the machine measurements, but if I got to choose from before to enter high quality measurements or after the winter lower quality measurements, I would always go for a lower cold after winter because that's when you have a lot of the wear on the roads if you have a wind in your country.
Yeah. No, that's that's a that's a very, you know, rich answer. There's a lot there. But I understand from what you're saying that your data is today used complementary to a lot of the existing more certified measurements. And the complementarity lies a lot in the fact that, of course you can get more, more measurements from you guys, more frequent, more daily measurements from you guys.
And at the same time, you can use this to single out any specific, let's say, bad quality segments where you want to deep dove and to have a more focused and more expensive, perhaps a measurement of the road.
Yeah. Or maybe even let me skip that step and identify those areas that are in between. I mean, that are I mean, you have the good roads and you have the failed roads and then you have the intermediate roads that might need attention or might be tangential. And I think that most people would want to go out and have a look on those roads anyways.
I mean, do a visual inspection. But it also comes down to what what level are you working on if you're working on a highway network or if you're working on a local city, road is low capacity roads tend to not be of scan with machine equipment. They usually scan a problem from visual inspections today so on. And there I think it's an easier way to replace it with an objective and more thorough measurements and from the D.O.T. perspective would be to highlight the areas where there's rapid changes or sudden damages of road at or actually for the winter.
We haven't done the certification, but one of the experts in my old university, Lulea Technical University, if you want to guess again, he's one of the guys working with the Winter Friction and has been doing so for a very long time. He has on the OR on the Swedish D.O.T. have bought the services from the university to evaluate how well we can use our data in comparison to those friction trailers.
And pretty much there's a white paper you can read about it, or maybe it's a Ph.D. I don't remember, but there's a paper you can read on the suitability of the technology we're using. And pretty much I mean, he gives or that paper that gives a thumbs up for technology we're using as a correct friction measurement device. I mean, the we measure the friction, correct.
But we can do it in a much, much wider network than the legacy machines or I mean, they could generate more measurements. But if you can afford to put like 200,000 of those friction measurement units behind trucks, that's that's that's the choice of of the vendor, of course, or the operator. But in practice, they're too expensive to be used throughout the entire road network, such as the the scanning equipment's being used on every single city road.
It's, it's too costly for all procedures in the road. So you don't get the reason whether.
Silver your thank you that's very useful to to here how does it compare with a more traditional and more certified measurements. Last I would like to ask you, you know, more generally, given the fact that you guys are really innovating on have a data project that's quite nascent and what are some of the challenges that you're facing in interacting with new consumers and getting your data to the market and then really making it used by the right people?
I think the last question or the last answer was a big part of it. I mean, with the certification procedure and those kinds of ways of working, that needs to be addressed. But besides that, I think it's also the people and the legacy of of organizations that are used to do it in one way, and then you're going to change it into some other way.
You work with the data and then you always get challenges. I mean, if you're changing, I mean, there's there's a reason there's change management consultants more or less that driving change in a business or even an organization takes time and it is a challenge. So I think that's something I mean, to to get get to the way where our tools really is intuitive is simple enough to be used by anyone in my my vision or our vision is that you shouldn't need to log into the system more or less.
You should just get a weekly update saying this is what you need to look on, and then you can go and have a look on those areas and also getting the reports of your general road network, the state of it. So kind of simplifying it to the way where it's almost too easy not to use it. I mean, that's where where we want to go with with the tools, especially for the asset management part of things.
But I don't think we're there yet and I don't think we know enough. But then I think we also need to try it out with having more of these processes where we run with with users real uses to have some of the main knowledge or through partnerships that we have with those construction companies. I mean, those companies, I think those are really valuable as well.
We're working with a bunch of those in in different countries, but getting the the acceptance to use our tools and also making it easy. I think that's the biggest challenge.
Got it. Yes. And I have to say that the demo that you saw this and that that spurred is one of the more user friendly ways of consuming this type of data. So I think this can definitely create more acceptance into the market than the oil. Streamlined your way with some of these clients. Great. We want to thank you very much.
This has been a super enlightening conversation. And I don't know if there's anything you want to add for anything we forgot to cover that you would like to add.
And no, I don't think well, there's a lot of things we didn't add, but I don't think there's anything that I, I must mention. So I think I'm happy with, with what the questions and answers will.
Have more opportunities to speak more. But yeah, super keen to build up our collaboration on the adoption of your data and thank you very much for today. Always a pleasure to collect that.
Thank you, George.
Super. Thank you.
Interested in learning more about Nira Dynamics' data products?
Book a free consultation with us and let's define your data needs together to achieve your objectives