Determination of Planetary Meteorology from Aerobot Flight Sensors

Abstract

Airborne robots, or aerobots, are fast becoming potential experiment delivery and planetary analysis vehicles. With the current trend towards a faster, better, cheaper methodology, aerobots seem to offer significant advantages over rovers and other methods of planetary exploration. This paper describes work performed in the Department of Computer Science at the University of Wales, Aberystwyth, as part of a PhD project, to examine potential methods that could be used to make an aerobot mission a realistic prospect. An essential part of any aerobot mission will be environmental analysis using a number of specialised sensors. These sensors are potentially made up of accelerometers and gyroscopes together with a wind sensor such as an anemometer and possibly temperature sensors. Such sensors will be used to allow the aerobot to quantify the weather patterns of the target planet and to determine the best course of action to achieve its goals. For a planet such as Mars, we have a relatively good knowledge of the global climate, but for an aer obot mission to succeed we must be able to analyse and determine the local climatic conditions around the aerobots' current position. In keeping with the faster, better, cheaper methodology it is important to minimise the number of sensors carried by an aerobot. It therefore seems reasonable to provide sensors that can play a dual role in order to maximise the use of the gathered data. That is, if environmental data could be extracted from sensors primarily used for flight control, we would be further minimising cost whilst not impacting the level of scientific analysis performed. The main focus of the work at Aberystwyth has been to investigate a potential method for retrieving environmental data from sensors that are primarily used to control flight. Using the FlightGear open source flight simulator together with terrain produced from MOLA data and realistic Martian weather patterns extracted from a Navier Stokes simulation, we have been able to produce a model of Mars that affords the ability to simulate an aerobot on a planetary exploration mission. In order for the aerobot to survive on Mars it must be able to react quickly to situations without operator input due to the long communications delay. Therefore any mission information sent to the aerobot from Earth must be at an