Such encounters are rather more common on the bustling roads of India, and whilst as a cyclist, it’s relatively easy to avoid such interlopers, especially if you’re crawling up a mountain, for human powered and autonomous cars, the challenge is a serious one.
A recent paper examines how such circumstances can be safely navigated, and indeed how the technology being developed for autonomous vehicles can be placed into existing cars as a driver aid.
The researchers, from the Department of Electronics & Communication, at Gujarat Technological University, in Ahmedabad, India, point out that livestock and other animals are a frequent presence on Indian roads, and contribute to a significant number of accidents.
India has the second largest road network in the world, and roughly 1 in 20,000 people die there in a road traffic accident. What’s more, 12 in 70,000 are seriously injured.
The team have developed a collision alert system using a dashboard mounted camera and a smart algorithm that can spot whether the objects detected as the car moves are on-road cows, and subsequently whether the movement of the cow presents a risk to the vehicle.
The system then alerts the driver via a timely audio or visual indicator to nudge them into applying the brakes. Suffice to say, the system is at a nascent stage of its development, and requires improvements to work as effectively in night-time or other poor visibility conditions,
It’s one of a number of projects that are designed to help detect more unusual road users. For instance, I wrote recently about a project designed to better detect cyclists on the road.
The algorithm they’ve developed forms a central part of what they call Deep3DBox, which is able to take a 2D image, identify road users within it, and then create a 3D box that surrounds each of them. It is also capable of determining the direction the vehicle is going in. When the algorithm was tested, it was able to correctly identify 89% of vehicles, but cyclists were another matter, with just 75% of bicycles identified, with much less success determining their direction.
Better scanning equipment and maps will no doubt help improve this statistic, but the inherent unpredictability of cyclists adds an extra layer of difficulty to the task.
“Bicycles are much less predictable than cars,” the authors say, “because it’s easier for them to make sudden turns or jump out of nowhere.”
Autonomous vehicles are already pretty good at detecting motor vehicles on the road, and it’s promising to see projects emerging that help them spot the various other forms of road user too.