Ford is Developing Autonomous Car Tech Designed to Combat Snow

Thinking Highways
By Thinking Highways March 9, 2016 14:54

Ford is Developing Autonomous Car Tech Designed to Combat Snow


Ford has revealed how the company will allow its driverless cars to cope with snowy conditions in a bid to accelerate mass-market appeal.

Most driverless car tests, ranging from those of technology companies like Google to traditional car makers like Mercedes and BMW, take place in controlled environments or on sun-drenched Californian roads.

However, Ford explained that driverless cars also need to cope with difficult situations such as wet, snowy and slippery surfaces.

“If self-driving cars are to become a reality, and they almost certainly will, they must be able to navigate snow-covered roads,” said the US car maker.

Ford’s response to such challenges has been to make use of Lidar technology, which useslasers to measure distances and creates 3D maps of the data during favourable weather conditions.

The maps are then used as a baseline of data from other driverless cars to identify their positions in situations where road visibility is poor.

Ford had to overcome the problem of Lidar sensors that were too sensitive and identified snowflakes and raindrops as obstacles. This returned false information to the autonomoussystem, giving the impression that precipitation should be avoided.

The company is now working with researchers from the University of Michigan to create an algorithm that can recognise snow and rain and filter out the information from Lidar sensor feedback.

Ford combines data from radar and cameras and mixes it with GPS and Lidar data to monitor the road environment around its autonomous cars. The sensors also scan the environment for landmarks, which are then contrasted against the 3D map information to provide what Ford claims is more accurate positioning than GPS.

All this information is then combined into a system that collects around 600GB of data an hour. Ford calls this sensor fusion.

The combination of data from multiple sensors means that a driverless car is not reliant on data from individual sensors. If one sensor malfunctions owing to deterioration from mud, ice or grime, for example, the car is not left without the necessary data readings.

Ford claimed that further advances in its technology may render this problem moot. “Eventually, the cars might be able to handle ice and grime build-up themselves through self-cleaning or defogging measures,” the firm said.

Ford may champion its driverless technology, but the company is not alone in testing driverless cars in difficult conditions. Swedish car firm Volvo has tested its autonomous systems in the snowy conditions of Gothenburg.

Thinking Highways
By Thinking Highways March 9, 2016 14:54