Researchers at Massachusetts Institute of Technology have developed an algorithm that will allow a plane to fly without GPS.
In response to the AUVSI’s AV challenge, which asked entrants to create a successful algorithm for indoor navigation by autonomous helicopter, MIT’s Robust Robotics group developed a more complex algorithm for the operation of an autonomous aeroplane.
“The reason that we switched from the helicopter to the fixed-wing vehicle is that the fixed-wing vehicle is a more complicated and interesting problem, but also that it has a much longer flight time,” said Nick Roy, head of the Robust Robotics Group.
“The helicopter is working very hard just to keep itself in the air, and we wanted to be able to fly longer distances for longer periods of time.”
The plane was built from scratch in MIT’s laboratory to a custom design and features a slow flying speed, with broad and short wings, affording tight turns and the cargo capacity to take onboard the necessary navigation and control electronics.
It successfully completed a maiden flight in an underground car park, weaving unaided through the building’s concrete pillars.
This flight marks a breakthrough for the AV arena, particularly in the field of aeroplanes, as a fixed-wing aircraft travels faster than a helicopter, and it is unable to undertake arbitrary motions, move sideways or hover.
The plane is loaded with a digital map of its environment and uses a laser rangefinder and inertial sensors to determine its location.
It also has to determine its orientation, velocity and acceleration, requiring it to calculate 15 different values in a complex multidimensional state.
It does this by combining two types of state-estimation algorithms, one accurate but time-consuming and another less accurate but more efficient.
Next on the researchers’ agenda is to develop an algorithm that will allow the plane to build a map independently as it flies.