Robots map with SLAM-dunk accuracy

Engineers at Purdue University, Indiana, have developed robots that use a variation of the Simultaneous Localisation and Mapping (SLAM) algorithm to make ‘educated guesses’ about what lies ahead as they travel in unfamiliar surroundings.

The new software algorithm, called P-SLAM, uses data collected from sensors, which include a laser rangefinder and odometer, on a robot to create partial maps as it travels through an environment for the first time. The robot then uses this partial map to predict the potential layout of an area.

According to Dr CS George Lee, professor of electrical and computer engineering at Purdue, the robot can navigate an environment more successfully the more repetitive the environment.

‘For example, it’s going to be easier to navigate a parking garage using this map because every floor is the same or very similar, and the same could be said for some office buildings,’ he said.

The Purdue researchers tested the algorithm in simulated and real robots in the corridors of a building on the Purdue campus. They found that a simulated robot was able to successfully navigate a virtual maze while exploring 33 per cent less of the environment than would normally be required.

‘Its effectiveness depends on the presence of repeated features, similar shapes and symmetric structures, such as straight walls, right-angle corners and a layout that contains similar rooms,’ said Dr Lee.

Future research will enable four robots to work as a team when exploring an unknown environment by sharing the mapped information through a wireless network. The scientists will also develop an ‘object-based prediction’ system to allow the robots to recognise objects such as doors and chairs.

The robots could be used domestically, or for military and law enforcement purposes to search buildings and other environments.