Google’s Self-driving Cars Learn to Read Cyclists’ Hand Signals

Since Google began testing autonomous, or self-driving, cars it has continuously updated the technology.

For example, last month the software named “Google Chauffeur” learned how to toot its horn, honking with appropriate intensity and frequency depending on the situation.

Now its autonomous cars “can detect a cyclist’s hand signals as an indication of an intention to make a turn or shift over,” according to the latest Google Self-Driving Car Project Monthly Report.

Beyond that, Chauffeur uses machine learning to remember signals from previous riders so it can better anticipate a biker’s turn down the road.

The report also says because the cars can see 360 degrees, they are more aware of cyclists on the road—even in the dark.

Through observing bicycles on roads and a private test track, the gumdrop-shaped self-drivers recognize many different types of bikes—from multicolored frames, big wheels, bikes with child seats, tandem bikes, and unicycles.

SEE ALSO: Tesla Model S Seen Driving With LIDAR Sensor on Roof

Most importantly, according to the report, “our cars won’t squeeze by when cyclists take the center of the lane, even if there’s technically enough space.”

Cyclists and drivers have a long history of not sharing the road, and even though the number of bicycle lanes has dramatically increased over the past few years across the country, more than 50,000 cyclists were injured and over 720 were killed in the US during 2014.

How autonomous vehicles will interact with cyclists and pedestrians remains a major barrier to the introduction of self-driving cars to the consumer market.

This latest Google report comes as the safety of self-driving cars is now being questioned.

A recent highly publicized crash revealed that a driver was killed while driving a Tesla with its semi-autonomous Autopilot mode in operation.

The National Highway Traffic Safety Administration (NHTSA) has begun a preliminary investigation into the incident and could possibly lead to a recall.