Ford develops technology that can predict human movement to reduce ‘petextrian’ collisions


Ford has conducted an intense program of more than 500,000 miles (804,670km) of drive testing in Europe, China and the USA, to develop a collision warning system that can help predict pedestrian movement.

Collisions can be caused not just by distracted driving, but increasingly, distracted walking. According to a recent report from the National Highway Traffic Safety Administration (NHTSA), a pedestrian is injured in a motor vehicle crash in the USA every eight minutes, a rate which is linked to a global influx of ‘petextrians’, people walking while texting.

Ford’s new Pre-Collision Assist with Pedestrian Detection can predict the movement of pedestrians to help reduce the severity of and, in some cases, eliminate frontal collisions altogether. Debuting in North America as an available technology for the 2017 Ford Fusion, the system has been developed to help recognize pedestrians using more than 240 terabytes of test data. Using 12 vehicles, Ford’s development team collected 473 days of data logging. In just 15 weeks in North America alone, the technology recorded more than three million scans of roadside objects, vehicles and pedestrians across roughly 70% non-highway and 30% highway miles.

The Pre-Collision Assist with Pedestrian Detection system uses combined radar and camera technology to scan the roadway ahead for collision risks. If one is detected, the initial response from the vehicle is to provide a visual and audible warning to the driver and temporarily mute the audio system. If the driver does not react to the warning, the technology can automatically apply up to the vehicle’s full braking force.

Working in daylight and in clear weather conditions at speeds up to 50mph (80km/h), the new system processes information collected from a windshield-mounted camera, which has been instructed to classify different vehicle and pedestrian scenarios, along with radar located near the bumper to pick up shape reflections. The technology then combines both data streams using a unique Ford algorithm that calculates the risk of a collision. Ford is now testing advanced features for future vehicles that could allow the system to operate at night, in low and harsh lighting conditions, and when vehicles and cyclists move in different directions.

“Since we are trying to avoid a collision that hasn’t yet happened, prediction of the future is an inherent part of the puzzle,” said Scott Lindstrom, Ford driver assist technologies manager. “Having a huge cache of data, based on real-world driving conditions, helps our system be smart enough to determine what may happen in a second that has not yet even occurred.”

Aaron Mills, Ford safety engineer, said, “The test data was gathered and then used to develop the vehicle’s algorithm to help it recognize a wide variety of human sizes and shapes. We were startled to see how oblivious people could be of a 4,000lb car coming toward them. It was a real eye-opener to how distracted people are today.”

Share this story:

About Author


Tom has edited Traffic Technology International (TTi) magazine and its Traffic Technology Today website since May 2014. During his time at the title, he has interviewed some of the top transportation chiefs at public agencies around the world as well as CEOs of leading multinationals and ground-breaking start-ups. Tom's earlier career saw him working on some the UK's leading consumer magazine titles. He has a law degree from the London School of Economics (LSE).