These notes are fragmentary but I will rework them.

First, various awards 

Main speaker: 

Discussions of the rovers on Mars

  • Discussion of the composition of Mars.
  • Robots sent to Mars
  • Software for driving not written before launch
  • Parachute.. Lander too heavy. Parachute redesign.
  • Airbags used as terminal velocity still 200kmh. Hard to design such that didn’t rip on rocks. Downwards pointing camera to check horizontal velocity
  • Petals of rover. Flat pack rover
  • Risk of rover falling over when leaving lander. Needed ramps
  • Programming movement due to time lag (at least ten mins each way). Vehicle needed to be somewhat autonomous. Computer 10MIPS so slow.
  • Working out driven path. Complex, wheel slip. Need to use vision. But when few features? Use own tracks, but make additional features when driving
  • Getting stuck. Use spare vehicles to test. Slip checks.
  • Dust devils. Dust devil recognition algorithm
  • 50Mb per Martian day transfer
  • Vision. Aegis. Edge detection, etc. Send minimum info to guide next day’s decisions.
  • Science. Layered rocks. Indicates there was once water on Mars.
  • Wheel failure at day 800 of the 90 day mission, but scratched surface and revealed science. That Mars was once, in places, wet and warm.
  • Reduced power output from dust, then hot blown clean.
  • Lasted 2000 days in the end. 

Opportunity rover

  • Looking for haemetite
  • Fell into a crater, but fortuitous as saw bedrock. Sedimentary rock. Little balls of haematite… Formed in water.
  • Rover lasted fourteen years. Killed by massive dust storm. 

Questions from audience 

  • Data: 15 mb engineering data per sol
  • Simulating Martian gravity? Use a lightweight version on rover
  • Had to use old tech that was very reliable and radiation proved, hence not the newest tech.
  • HPC to simulate vehicle? Thermal, FEA. CFD to model in atmosphere flight, but not the parachute.