One of the benefits of autonomous cars is that they can be more observant than humans. For example, they don’t get tired, aren’t affected by drugs or drink and can make decision faster than humans in most cases.
But what if they could also see round corners? If there was an obstacle in one of the roads it was planning to turn down, it could instead continue driving and find another route. If the traffic lights were red, it could instead find a quicker way of getting to your destination.
A team of researchers at Stanford University have developed exactly this technology, using lasers to predict what’s ahead, even if it’s not directly in front of the car. It’s based upon Lidar, which is the type of laser already used by autonomous cars to detect obstructions and hazards.
The method was tested using a a laser and photon detector. The two technologies were placed in front of a wall next to an object, which was separated from the source by a partition wall. The lasers were then fired at the first wall at an angle and the researchers were able to collect the data generated when the lasers hit the object and scattered.
This could then be processed using an advanced algorithm to create an image of the object.
“It is a very simple tweak to how you do imaging, but it has major implications in terms of how you can reconstruct the images from that information,” Dr Matthew O’Toole, a coauthor of the research from Stanford University said, explaining the method uses less memory than traditional techniques and it can also create a higher-resolution image too.
However, it’s far from perfect at the moment because scanning the original wall can take too long to make a split second decision, the researchers explained.
