How Tom Tom achieved its digital transformation with Big Data
Big data is more than ever on the agenda of those marketers who are on the warpath of data-driven marketing. It’s the 6th year I’ve been active (from a content marketing perspective) in this area and I find it always more exciting, year after year. On March 12, I attended the Big Data Paris 2019 keynote entitled “How Tom Tom has evolved from a navigation company to a big data company”.
Tom Tom lives and breathes Big Data
The speaker was Alain de Taeye, Founder of TeleAtlas, Member of the Management Board, Tom Tom. For those who wouldn’t know, Tom Tom is a Dutch company. His pitch was instrumental in my understanding how Big Data moved from a technical topic into a full fledged transformational toolbox reshaping entire industries and businesses.
Such was the promise of Big Data six years ago or more, but few are able to show such impressive results as Tom Tom. Here is how they turned from a B2C company selling navigation devices into a B2B data-driven money machine.
The Big data revolution was the subtitle of de Taeye’s presentation and God knows this was a revolution! Tom Tom has changed radically. Here is his account thanks to my notes taken during his presentation.
Making navigation easier with Big Data
“At this moment in Paris we are collecting data from users and this is used to make navigation easier” Mr de Taeye said as an introduction to his speech. Tom Tom is well known for its maps and rapidly evolved into a technology company. “The company’s maps aren’t ‘ordinary maps’” however, Mr de Taeye went on.
They are maps sent to the dashboards of your cars but they are also HD maps used in autonomous driving, as well as driving maps used by onboard computers. Tom Tom also offers APIs used by developers to connect to their maps and companies using their products like Microsoft (Azure is resorting to their maps for which they are a prime deliverer).
In essence, Tom Tom went from a B2C company selling navigation tools to a B2B company empowering the self-driving revolution of the future.
De Taeye is adamant “in the foreseeable future, not the next two years that is, more like the next 15 or 20 years, we’ll all ride in self-driving cars and those cars will all need different maps”. That’s a vision which I find rather optimistic but you never know with technology. Innovation often turns out to be a self-fulfilling prophecy and in the automotive industry it’s people like Tom Tom who shape the future. Maybe the important word in his sentence is “foreseeable”.
The question is “why big data”? 3 reasons why it’s also useful to other industries
There are three reasons why Big Data matters for all industries according to Mr de Taeye:
- Number one is problem solving: all complex problems need lots of data to solve. Mobility problems in Paris are huge, we all know that, and Paris isn’t alone. “They can only be solved with big data” declared Mr de Taeye,
- Number two is that real-time data has become mandatory. The assumption is that all info coming from the smartphone is in real-time and correct. This means “we have to have a perfect view of reality and it is in essence very complex and requires a lot of data”,
- Number three is Robotics: that is to say, performing tasks very similar to what human beings are doing. In self-driving cars we need big data to make the vehicle as safe as if it were driven by a human being.
How big is Tom Tom ‘s big data?
Quantifying Big Data is also a good idea. De Taeye gave us hints as to how much data the Dutch company is crunching. And the numbers are huge:
Tom Tom is teaming 19 trillion anonymous data points in total and all this amounts to 2.1 billion km of raw data which they collect on a daily basis. it’s an enormous amount of data but what’s more interesting is what can be done with it.
Here are some of their use cases.
Use cases of Big Data implemented by Tom Tom
- First and foremost, cities all over the world are encountering huge mobility problems. They all have to face massive infrastructure challenges and all these assets need to be mapped properly,
- Secondly, Big Data makes it possible for the authorities to access a real-time view of the use of this infrastructure including real-time traffic information and its historical usage data. All this makes it possible for them to manage traffic in real-time. All major cities are doing a pretty bad job in this area at the moment. “Imagine that traffic management in real-time be enforced properly”, warned Alain de Taeye, “traffic congestion could be reduced by 20% to 40%”. Here is an opportunity that big cities should definitely take into account,
- Thirdly, there are cutting edge use cases in the automotive industry. This sector is undergoing a Big Data revolution. Cars are becoming “computers on wheels” de Taeye explained. E-vehicles are different from standard ones and require data to keep going. Whether there is traffic or not, the range of an electrical vehicle will vary greatly and vehicle management will be adapting to the current situation too. Here we have concepts such as ADAS. An acronym that stands for Advanced Driver Assistance System: a contraption aimed at helping drivers reduce consumption of their vehicles. There are advanced ADAS cases like engine management: while driving on steep hills for instance, consumption rises and management changes depending on terrain. “Mostly if you drive a 20 ton truck” de Taeye pointed out,
- Fourthly there are the new forms of transportation such as Uber et al. All those want to do as much transport possible while reducing their ecological footprint, not to mention labour costs,
- There is a fifth business case, when drivers need assistance in finding parking spots. Most public car parks are now equipped with electronic systems which can send data back to the driver and street parking remote management is also possible. Into the bargain, traffic lights can be managed in a much better way,
- Lastly, there are also cases for which Big Data is absolutely mandatory like real time maps. “Real time maps are the holy grail of the automative industry” Alain de Taeye said. To that end changes must be taken into account in real time when they happen. “It’s no pie in the sky” de Taeye added, “it’s possible and we’ll be there in a little time”. To do this, Tom Tom personnel survey roads by foot and bike and car. For sure, “Google also uses vans to survey the streets” de Taeye says but they are the “same vans that Tom Tom has been using since 1989”; way before Google came into that space. Real-time maps are not a piece of cake however. “If you want to create real-time maps, manual work must be avoided” de Taeye explains. One must minimise the time required to change a map and refresh it. There is no more time left for surveying and all must be done on the fly. This is why Big data is pulled down from various sources such as satellites. All the data is then combined and matched with community-driven data. That’s not all. Production tools guaranteeing the quality of these systems are also required. “It took us five years to set up such a system and it has now been in operation since 2016” the Tom Tom exec said. Changes are tremendous and scalability is not an option. Such a system hinges on 2 million changes per hour: “that’s half a million database changes during the current presentation” de Taeye added.
Tom Tom focuses on self-driving cars
Why is the automotive industry so focused on self-driven cars? de Taeye reminded us that there are close to 1.3 million deaths on the roads and most of them are finding their cause in human behaviour be it alcohol consumption or other reasons. “One must prevent these errors from happening. 90% of it can be suppressed” de Taeye added.
For self-driving cars to work properly, maps need to have ultimate accuracy. Precision must go from 10 m to 10 cm, which means that totally new maps are required. Such maps must be absolutely perfect. If there are mistakes in maps used by humans, they all find that annoying. With robots it becomes unacceptable.
This means Tom Tom needs real time HD maps for self-driving cars to be working properly. Their creation should also be automated. Manual real-time HD map design would not just cost a fortune, it would be impossible. As seen above, Tom Tom personnel drives around with mobile-mapping driving vans. On this basis they combine the collected data with AI and produce HD maps almost instantaneously. Real-time map maintenance is also a must-have. Modern cars are equipped with radars and also cameras which can have a bird’s-eye view of the car while driving. “We’ve done experiments and proven that based on these sensors we are able to update the data in real-time and upload it, process it in real-time and send it back to the car in no time”. A real technological prowess.
One can imagine the difficulties because in the car there is an enormous amount of data available. In no way so much data can be sent to the cloud. Hence Tom Tom had to design a system whereby process of some of this data takes place in the car before it is sent to the cloud.
However impressive all this may seem, this is only the beginning and in 2-5 years from now, Tom Tom and the industry will be doing a lot better and the amount of data handled will be increasing exponentially. I’m sure we’ll hear about the progress made by Tom Tom and the industry at one of the forthcoming editions of Big Data Paris. Who knows, maybe we’ll get to that conference in a self-driving Uber cab using Tom Tom automatic navigation?!