Sergii Sviatokha, Automotive Practice Leader at Intellias talks about smart solutions for cars developed in Lviv.
Nowadays, sitting behind the wheel, almost all of us use a navigation system that helps us avoid traffic jams and plan the fastest route from one place to another. As technology advances, cars become smarter. The first self-driving cars are already on the road. In 10 years, you won’t even need to keep your hands on the wheel and keep focused on the road, because the car will do it by itself.
Being Ukraine’s hottest tech hub, it doesn’t come as a surprise that Lviv houses automotive company Intellias, which develops navigation systems, integrated map solutions, object recognition on the road, and Computer Vision. We talked to Sergii Sviatokha, Automotive Practice Leader at Intellias, about the company’s latest projects.
One of the most interesting cases we are currently working on is car-to-car communication. It means a car can recognize a dangerous situation, like a slippery road ahead, while driving. The car sends this information to the Cloud, which automatically transmits it to the cars behind you. Thus, these drivers receive a warning sign on their screen or are advised to slow down. Road conditions are constantly changing – slippery roads, traffic jams, accidents, sudden fog or construction work, so it’s important to continuously identify, analyze and distribute this info to cars approaching a certain road section.
The hottest topic in the automotive industry right now are cars that analyze and recognize a pedestrian’s behavior on the road. It’s a system of algorithms that register if a person walks aggressively or not, fast or slow, etc. This research attracts a lot of investments in the industry.
Modern cars are already equipped with various radars and sensors that reproduce a 3D model which shows what is in front, behind or close to you. Imagine, you are turning right, but your visibility is limited because of a tall building on the corner of the street. In this case, your car communicates with surrounding cars on the crossroads. Other cars in your blind spot transmit data from their point of view, completing your visibility information so you can make a safe turn. On top of that, cars can decide which one should go first or wait and give way to another car.
Driverless cars still amaze us today, but soon they will be part of our children’s everyday life. Even for me, it’s still hard to imagine that I will hand over control of my car completely to a computer. It’s not just difficult to relinquish control over the vehicle – for many people driving is a culture, hobby, or even life. Moreover, being a driver calls for a big responsibility. Most accidents are still caused by the human factor. If self-driving cars can reduce the number of accidents by at least 7% – it is worth it.
This new technology does not only apply to electric cars. Autopilots and the latest developments are already being used in petrol cars. Right now, only premium class cars have these features. For example, the new Audi 8, which was presented in July 2017, already has an autopilot. The car is equipped with four cameras with a circular view, one front camera, one infrared camera, a laser scanner and a medium-range radar. All these devices assist both driver and autopilot to drive more safely.
Car manufacturers are now working on an internal ecosystem for the car. A passenger of the self-driving car will have something to do during the trip. Microsoft already started integrating Office programs into Nissan or BMW cars. The car will offer a fully equipped environment for a passenger to work during the ride.
Taking away the human factor
Recently, Intellias launched a Computer Vision lab, where the team works on object recognition, like other cars, pedestrians, and road signs, as well as the analysis of weather conditions, and, most interestingly, the human behavior on the road. It’s complicated. For example, you are driving and a person starts crossing the road. You, as a driver, notice this person and he/she also looks at you, establishing visual contact. You understand whether the pedestrian will wait and let you pass or vice versa. A car could not establish such a connection on its own.
Moral dilemmas also pose problems which are still hard to solve on a technical level. For example, you are driving with your family in the car at high speed when suddenly a person starts crossing the road. In order to avoid deadly consequences, you need to make a choice – turn left and crash into a car in the opposite lane, hit the pedestrian or crash into a concrete building on the right. Which one do you pick? Would you risk killing the crossing pedestrian, the other driver or your own family? There is no right answer in this situation, which makes it nearly impossible to program it for a vehicle.
One of our recent projects is the development of integrated route and charge planning for electric cars. Route concepts are calculated to satisfy the needs of both the driver and the car. For example, when a driver wants to have some coffee or a snack, and when the car has to be recharged. The car offers the best route, taking into account all these details. We use Machine Learning for such complex calculations. At Intellias, we use systems capable of collecting a huge amount of data. We gather data about traffic situations, weather conditions, from high-precision maps, and from other cars through the connected car system. This data is processed by the computer, and based on that, the system makes the optimal decision.
The German government plans to get rid of traffic lights and road signs in the future. Thanks to the large amounts of data, high-precision maps, sensors and artificial intelligence, the cars will be able to navigate without any road signs or traffic lights. By exchanging data between cars, your car will find a place to park, know when and where to turn or stop, and when to wait and give way to another car or pedestrian.
Car as a service
According to Volkswagen, 95% of car owners only use their car to drive to work and back. The rest of the time the car is not in use. A future solution would be to utilize cars as a service. A car should become a service, not just a property – like Uber, Gett, or Yandex Taxi.
Volvo has recently launched a car service called “Care by Volvo”. It’s a monthly subscription for the new XC40. The cost of the subscription ranges from $500 to $1000 per month. You can order the car via a mobile app, and it shows up exactly when and where you need it. The price includes services, taxes, parking, etc. On the weekend, the user can choose another car, a bigger one for example. The service will be available in the US, UK, Sweden, Germany, Spain, Italy, Norway and Poland from 2018. I believe this is a good alternative for those who don’t drive that often.
Cities suffer under the amount of cars. This needs to be solved somehow. For example, in Kyiv, people spend up to 2 hours a day in traffic jams. If you cut the number of cars on the roads – there will be less traffic and CO2 emissions, and people will still get to work or other places every day.
A possible solution is Shared Mobility, a transporting service for several passengers traveling from one location to another. Carsharing services like Blablacar also work according to this principle: people drive their own car and pick up other passengers on the way. This way of mobility will become even more popular with the rise of the driverless cars.
Even though the automotive industry is developing at a high rate, the implementation of the driverless ecosystem requires the approval and assistance of countries and their respective governments. For example, Norway plans to completely ban petrol powered cars by 2025. Owners of electric cars already benefit in the country: they have fixed parking spots, pay lower taxes, etc. Such advantages influence consumer behavior. The country has the largest number of electric cars in the world and this year they set a sale record as Volkswagen delivers 996 all-electric VW e-Golfs.
Already back in 2014, the European Parliament agreed to reduce greenhouse gas emissions by 40% by 2030. Volvo plans to stop producing petrol and diesel cars before 2019, replacing all models with electric and hybrid. Germany’s federal council, the Bundesrat, has passed a resolution calling for a ban on combustion engine cars by 2030. The UK will ban sales of new petrol and diesel cars starting in 2040 as part of a bid to clean up the country’s air. By 2050, all cars on the road will need to have zero emissions.