DETROIT MOTOR SHOW: Ford tests autonomous cars in snow

Ford has claimed an industry first, testing its autonomous Fusion in snowy Michigan.

The manufacturer says the combination of high-resolution 3D mapping and LiDAR can overcome the lack of visible line markings or gutters at the side of roads. It is this breakthrough that has paved the way for successful testing of the self-driving Fusion (known as the Mondeo in Australia) on snow at Ford's Mcity test facility.

"It's one thing for a car to drive itself in perfect weather," said Jim McBride, Ford technical leader for autonomous vehicles. "It's quite another to do so when the car's sensors can't see the road because it's covered in snow.

"Weather isn't perfect, and that's why we're testing autonomous vehicles in wintry conditions – for the roughly 70 percent of US residents who live in snowy regions."

According to Ford, autonomous cars can't rely on GPS and line markings alone to navigate their way around. The global positioning system isn't sufficiently accurate to locate the car precisely. Ford says that positioning a vehicle within its lane requires accuracy measured to within centimetres. About as close as GPS can come to that is a margin measured in metres.

That explains the importance of LiDAR in Ford's system. The system emits laser light pulses that bounce back (as RADAR systems do with radio transmissions) to create a 3D image of the driving environment. It is much more accurate than GPS, Ford says.

Unfortunately, LiDAR is only as good as the prevailing environmental conditions allow it to be. The laser light won't penetrate snow to single out kerbing, for instance. And sensors receiving the light impulses can be occluded by road grime.

That's the point where the car's 3D mapping comes into play. Assistance provided by researchers at the University of Michigan ensures the mapping data is interactive and the autonomous system effectively learns different drive routes and the location/position of road infrastructure in favourable conditions. The map data then serves to guide the car when those points of infrastructure are not visible – beneath a layer of snow.

"Maps developed by other companies don't always work in snow-covered landscapes," said Ryan Eustice, associate professor at University of Michigan college of engineering.

"The maps we created with Ford contain useful information about the 3D environment around the car, allowing the vehicle to localize even with a blanket of snow covering the ground."

Eventually, Ford anticipates that stability and traction control systems will interact with autonomous driving systems to determine whether it's safe to proceed.

"The vehicle's normal safety systems, like electronic stability control and traction control, which often are used on slippery winter roads, work in unison with the autonomous driving software," said McBride.

"We eventually want our autonomous vehicles to detect deteriorating conditions, decide whether it's safe to keep driving, and if so, for how long."

Since the middle of 2015 Ford has made the transition from the first (research) phase to the second (advanced engineering) phase of developing autonomous motoring for production. The manufacturer has tripled its autonomous motoring test fleet to 30 vehicles for assessment in California, Arizona and Michigan. Each vehicle is equipped with LiDAR capable of operating at a range of up to 200 metres.