Self-driving cars have been a part of the conversation in the tech world for years now. Whether it be conjecture in science fiction from decades past or some of the real progress we’ve made recently, humanity has had a long-standing love affair with the concept of totally automated transportation that we’re only now starting to be able to make into a reality – although there continue to be several bumps along the way. Enter the IEEE Sensors Journal, which has just published an article that leads us to the conclusion that the technology behind the Kinect sensor could be the key to getting past some of the roadblocks that have kept us away from self-driving cars.
You can read the abstract of the article right here, and if you’re willing to spend a few bucks, you can read the entirety of the article. The long and short of it is that the Kinect sensor’s technology is able to overcome some of the shortcomings of the sensors that are currently in use when working with self-driving cars. The self-driving cars that run via a connection to cameras using stereovision, and while this technique is generally pretty functional, it’s also vulnerable to “solar radiation contamination,” which makes the projection of light unreliable, and could potentially cause a self-driving car to make some errors while on the road – potentially fatal ones.
In its attempt to fix this problem, the Kinect endeavors to “translate the depth image provided by the sensor into definite obstacle projections in the navigability map used by the vehicle.” In other words, the Kinect uses its many layers of detection to create a more comprehensive and reliable map of obstacles for the vehicle that it’s hooked up to.
While this is just an article and not anything being put into effect by a major company anytime soon, the concept here is solid, and if taken into account by the engineers working on self-driving cars, may result in a safer experience when these vehicles hit the streets.Further reading: Kinect, Microsoft, Self-driving cars