A robot can autonomously navigate inside a building using nothing but a magnetometer and a detailed map of local magnetic anomalies.
The technique could provide a means for people and robots to find their way around large buildings, but the technology may be some way off commercial application because of the hefty cost and the size of the sensors.
Satellite navigation systems like Russia鈥檚 GLONASS, the European Union鈥檚 Galileo and China鈥檚 BeiDou can provide accurate location information all over the planet, but they do have weaknesses. Signals can be jammed in times of conflict, as , and they tend to be difficult to pick up indoors.
Advertisement
The US Air Force has run tests of magnetic anomaly navigation on fighter jets, which relies on the unique patterns produced by the geology of Earth鈥檚 crust and, crucially, can鈥檛 be jammed. Now, at the University of Florida and his colleagues have collaborated with a scientist from the Air Force Research Lab at Eglin Air Force Base in Florida to demonstrate that a similar technique can be used indoors by autonomous robots.
The team created a detailed map of the magnetic anomalies in an 11-metre-by-6-metre area of a laboratory and showed that a robot equipped with a magnetometer could navigate to specific waypoints despite not being told its starting position.
Ganesh says Air Force tests have tended to focus on taking magnetic measurements and GPS measurements together to test the accuracy of the concept, and that his team鈥檚 robot is one of the first demonstrations of real-time, autonomous magnetic anomaly navigation.
The team鈥檚 small robot takes a reading of the magnetic field at its initial location and generates a database of possible locations based on its detailed anomaly map. As the robot moves 鈥� at a constant rate of 20 centimetres a second 鈥� it gathers more readings, whittles down the guesses and ranks them in terms of probable correctness. Within several readings, the robot can determine where it is on its known magnetic map and begin to navigate to its target destination.
Ganesh says that when mapping their laboratory, the sensors picked up on metal pipes that lay underneath the floor, which became landmarks for the robot just as mountains or ridges would for a fast-moving aircraft.
鈥淚t works in small and large scale,鈥� he says. 鈥淪o, there鈥檚 loads of similarities between what the US Air Force uses and what we鈥檙e doing. Magnetic navigation is available everywhere.鈥�
Although the team has proved that magnetic mapping could be a useful tool for autonomous robots and humans alike, it may be some time before it becomes a practical consumer tool because of cost. The magnetic sensors used in the experiments cost $10,000 and, at 16 centimetres long, are too bulky to fit inside a smartphone or tablet.
Reference:
