NIA | AuRA Lecture Series | Jean de Lafontaine

AuRA Lecture by Jean de Lafontaine

Date: October 5, 2006
Time: 1:00pm ** NOTE NEW TIME **
Location: NIA, Rm 137

Autonomous Hazard Avoidance and Precision Landing on Planetary Bodies
Jean de Lafontaine, NGC Aerospace Ltd.
In the search for a better understanding of our solar system, remote sensing of planets from orbits and in-situ analysis of their surface have provided researchers with invaluable scientific data. Future missions will aim at the return of planetary samples for analysis in Earth laboratories. This quest for better knowledge requires the landing of space vehicles in scientifically meaningful but often hazardous regions, hence the need to detect and avoid surface hazards. Pin-pointing landing accuracy is also required in order to reach desired regions and minimise the mobility requirements of surface mobile robots. For the past 5 years, under contracts with the Canadian Space Agency and the European Space Agency, NGC Aerospace Ltd has developed and validated innovative guidance, navigation and control (GNC) algorithms to achieve autonomous hazard avoidance and precision landing on planetary bodies. In most cases, a Mars landing mission provided the reference scenario but some studies addressed the safe landing on atmosphere-free bodies (Moon, Mercury) and asteroids.

Inspired by the work of A.E. Johnson and colleagues, the hazard-avoidance techniques utilize the concept of topographic cost maps, a geographical representation of the terrain where the altitude is replaced with a cost function proportional to the risk of landing at given site. This risk is a function of the surface roughness and its slope relative to gravity. An extension of this concept, the so-called fuel cost maps, specifies in addition the propellant required to reach each candidate landing site. The safest and easiest-to-reach landing site, so identified by these cost maps, is fed to a guidance and control system based on a Viking-type gravity-turn guidance law and an Apollo-type quartic guidance law. This closed-loop autonomous system thus ensures a safe landing at a safe site. Contrary to hazard avoidance, which is active during the last few kilometres before touch down, the precision-landing techniques for Mars involved the development of robust guidance and control algorithms from atmospheric entry at 120 km through the hypersonic braking phase, the parachute phase and the propulsive descent phase. The main innovations in this last phase include velocity-determination algorithms based on pseudo-Doppler and feature tracking techniques and terrain-navigation algorithms based on feature recognition. In both cases, one of the most promising techniques consists in treating surface features as if they were "stars" and use the flight-proven constellation-recognition algorithms of autonomous star sensors to recognise these surface "constellations". Although all the developments and laboratory tests conducted so far by NGC relied on three-dimensional data obtained from a Lidar mapper, further developments have been proposed to extend these techniques to camera sensor coupled with an optical correlator and an optical flow algorithm.

The presentation will provide an overview of the Lidar-based hazard-avoidance, velocity-determination and terrain-navigation techniques as well as a description of the Autonomous Planetary Landing Simulator (APLSTM) used for their validation. Monte Carlo simulations and a video will illustrate the performance of the closed-loop GNC algorithms. Static laboratory tests conducted with real Lidar scans of an emulated Mars surface will also demonstrate the quality of the cost maps and the computational load required to generate them. The presentation will conclude with a description of the dynamic test facility under construction at NGC's premises and an overview of the helicopter-based flight tests under preparation, both aiming at the achievement of Technology Readiness Level 6 within the next couple of years.

Bio: Jean graduated in engineering physics at the Royal Military College (Kingston, Canada) and obtained his PhD in aerospace engineering at the University of Toronto. From 1982 to 1986, he worked for the Canadian Government on the development of a military satellite project. From 1986 to 1996, he was employed with the European Space Agency (ESA) as a satellite system engineer both at ESA and at the NASDA Tsukuba Space Centre near Tokyo. Jean returned to Canada in 1996, became a professor at the electrical and computer engineering department of the Université de Sherbrooke and founded his company NGC Aerospace Ltd.

As the president of his flourishing company NGC Aerospace Ltd, Jean currently manages a number of space projects with the European Space Agency, the Canadian Space Agency and various aerospace companies. He also continues to teach control system theory as a professor of electrical and computer engineering at the Université de Sherbrooke. One his most recent achievements consists in the development of the PROBA-1 flight code, the first fully autonomous guidance, navigation and control software to be flown by the European Space Agency. PROBA-1, an Earth-observation mini-satellite, was launched in October 2001, initially for a two-year mission. It is still operating today with flawless performance.