The Stack Archive

Research proposes organised chaos for driverless traffic intersections

Mon 5 Sep 2016

Researchers are developing smart intersection systems for driverless vehicles operating in complex traffic situations, such as those found in many Asian cities. With motorbikes, cars, and pedestrians all travelling independently and often with relative disregard for others, it is difficult to imagine a self-driving vehicle successfully navigating the chaos.

Now, a research team at Carnegie Mellon University, led by professor Raj Rajkumar, is attempting to build a driverless solution to the traffic idiosyncrasies which autonomous technologies face in places such as Vietnam, China, and India.

Rajkumar said that it will take 10 years or so to develop an effective system. ‘Sensors on self-driving cars will have some inaccuracies to deal with, and also will require a good amount of computing power. Designing, implementing and testing these will take time,’ he said.

The professor posited that the industry must work together to ‘decipher’ the way independent traffic situations roll out in Asian countries, compared to Western models where drivers obey established road conduct rules.

Rajkumar explained that in Asian traffic systems ‘there is constant monitoring, adaptation, and reaction.’ He added that ‘everybody subconsciously knows how long it would take to stop oneself (be it a vehicle, cart or human) etc. to avoid an accident. Beeps and horns play a major part. Eye contact makes a difference. All parts interact and play a role.’

These country-specific styles need to be interpreted and encoded into driverless systems. Rajkumar suggested that the best way to go about this is to develop ‘vehicle to vehicle’ and ‘vehicle to pedestrian’ communication – technology which can connect all people and vehicles automatically.

Carnegie Mellon PhD graduate and autonomous driving research scientist at Ford Motor Company Reza Azimi posted the following video depicting a driverless solution to intersection traffic:

However, achieving a successful system like this presents a huge challenge, argued Rajkumar, with every car and individual involved needing to be equipped with the correct technology. He suggested that integrating the technology into mobile phones would be a useful method for diffusion – particularly if regulation was updated to support the spread of the technology.


Asia news research self-driving cars
Send us a correction about this article Send us a news tip