Autonomous systems developed for ESA’s ExoMars rover, which will allow it to analyse Martian terrain and identify the best point on rocks to drill for samples without need for human intervention, could treble the speed at which the rover can collect a sample, compared to previous Mars rovers.
In simulations, now being backed up by laboratory tests in the ‘Mars Yard’ at the University of Wales, Aberystwyth, the rover first builds up a three-dimensional model of its surroundings and then analyses each rock for surfaces suitable for drilling. The rover can then calculate the adjustments needed to position its chassis, robotic arm and instruments to acquire the sample.
‘This system allows the rover to do more than find flat areas to drill. The versatility of our system and its ability to pinpoint the best site to take samples, even from complex micro-features on rocks, could be vital when looking for evidence of exobiology,’ said Dr Dave Barnes, a Reader in Space and Planetary Robotics at the Computer Science Department at the University of Wales, Aberystwyth.
In recent Mars missions, up to 40 per cent of operations time has been taken up with defining, planning, rehearsing, scheduling and uploading every move that the rover makes on the surface of Mars. For NASA’s Mars Exploration Rovers, three Martian days can elapse between a target being identified and the rover actually acquiring the sample. The autonomous systems developed by the Aberystwyth team should bring that time for ExoMars down to less than one Martian day.
Software developed by the team, who worked with EADS Astrium on the Phase A study for ExoMars, uses stereo images to build up a digital elevation model and to classify features into six categories: peaks, ridges, passes, planes, channels and pits. The level of detail for each feature can be varied by adjusting the number or data points, the slope and the minimum curvature for the model. The rover selects a suitable surface, then ‘tags’ the optimum drilling point and calculates how to move the instruments at the end of its robotic arm into position.
‘We are now starting an exciting experimental phase of study with our Concept-E rover chassis model, which has six wheels that can drive, turn and move up and down independently. This gives us eighteen degrees of freedom when adjusting the pitch, roll and yaw of the chassis. We are working on a unified control system for the chassis and the robotic arm, which itself has four degrees of freedom, so the rover can manoeuvre itself to access samples even in hard to reach places. This puts us at a new level of manoeuvrability compared to Mars landers that have flown to date,’ Dr Barnes said.
The Concept-E rover will be operated on the newly completed Planetary Analogue Terrain (PAT) at Aberystwyth, a 50 metre squared sculpted landscape, complete with a drilling pit, covered with soil and rocks that have been selected for their Mars-like properties.
‘The majority of our work to date has been in simulation but there is no substitute for experiments with real hardware. We are looking forward to repeating our experiments with a real rover and instruments in our new PAT laboratory,’ Dr Barnes said.
ExoMars, which is scheduled to launch in 2013, is the first mission in the European Space Agency’s Aurora programme to explore Mars and the Moon. It will search for traces of past and present life on Mars and gather information the Martian environment in preparation for future missions. The ExoMars rover will carry a comprehensive suite of instruments dedicated to exobiology research. The rover will be able to travel several kilometres during its nominal lifetime of six months and analyse samples from with in surface rocks and from the subsurface, down to a depth of two metres.