US researchers from Purdue University, Indiana are developing robots that navigate through unfamiliar surroundings by creating maps of areas they have already explored and then use this information to make predictions. A technique called "simultaneous localisation and mapping", whereby the robots build up a map of their unknown environment using various sensors whilst keeping track of their current position was labelled as slow and prone to errors. Instead, the robots create a 2D map of the area they are exploring, but when they come to an unknown area, they use an algorithm to check back through the stored information to see if it seems similar to any areas that have already explored. They do not bother exploring areas they conclude are similar to the ones they've documented.
The scientists found that using this technique, the robots could navigate successfully while exploring a third less of their environment and were less prone to errors than robots that used SLAM. Professor George Lee, who carried out the research, still mentioned limitations: "Indoors, in places like office buildings, it works well; outdoors, where the scene isn't as repetitive, the result is not that good." Professor Lee's research was funded by the National Science Foundation and was published in the journal IEEE Transactions on Robotics.
News source: BBC News