It's one thing for an autonomous car to strut on a warm, warm Californian tarmac, and another to do it on the frozen winter mix of northern Finland. Martti, a self-guided autonomous vehicle system in Finland, demonstrated that during a record drive along a treacherous Lapland road (to normal drivers).
Martti is one of two cars designed by the VTT Technical Research Center; it is designed to handle harsh and icy conditions, while its "spouse" Marilyn is made for more ordinary urban journeys. Different situations require different sensors and strategies – for example, simple optical cameras work poorly on snow-covered roads, and lidar is less efficient, so Martti will rely more on the radar. But Marilyn has a lidar mounted on the back for a better awareness of the situation in traffic.
Recently, Martti accomplished what the researchers claim to be a world first: driving fully autonomously on a real snowy road (and reaching 25 mph). Others, from Yandex to Waymo, have tested cars in the snow, but according to their reports, these conditions appear to have been more controlled. The Martti promenade was held in Muonio on a public road almost totally obscured by the snow.
" He probably also made a new world record in fully automated driving, making 40 km / h in snow on snow-covered terrain with no lane markings," project manager Matti Kutila said in a report. MTB release. "He could have had even more speed, but in test, he is programmed not to exceed the limit of 40 km / h."
I'm not sure that going faster would be wise even on straight lines. But driving in the winter is not my specialty.
It is not a question of making a perfect consumption car for snowy weather, but of tackling the unique technical problems that come with it. For example, can Martti's optical instruments adjust the wavelength that they use based on the conditions in order to obtain a small but significant increase in accuracy? What about the detection of icy conditions and traction problems in advance – how should the car collect this data, and how should it act until the We know what to do?
Inter-car networks can be critical for this, suggest the researchers, including other autonomous cars on the road and specialized vehicles that can test and broadcast information such as snow, traction, road temperature etc.
In the end, this intelligence could prove very useful for applications such as automatic tractors, forestry trucks or emergency vehicles.
Soon the team will go even further in the Nordic environment: " Next spring, one of our vehicles can also be spotted in forest environments, when Marilyn and Martti have a new friend able to attack on all terrains. "
Image of hotel: ATV