The recent 16th International Mars Society Convention at University of Colorado, Boulder, comprised 4 days of packed plenary sessions, outreach programs and presentations from an eclectic crowd of enthusiastic experts, students and amateurs. NASA scientists and engineers described their work on Mars landers; students presented results from simulated off-world greenhouses; private sector entrepreneurs promised the exploitation of asteroid resources and the founding of future Mars colonies; and I presented The University Of Manchester’s pattern recognition system for making measurements from planetary images.
The University Memorial Center (with the American ‘er’) played host to around 400 attendees during the event, with over a dozen invited keynote speakers presenting over the course of 4 mornings, and dozens more individuals presenting across 3 parallel track sessions each afternoon. The range of topics spanned disciplines, from engineering to life sciences, from radiation monitoring to reusable heat-shields, and from composting with worms to corridor discussions about time travel through wormholes.
Hosted by Dr Robert Zubrin, founder of the society and author of ‘The Case for Mars’, the conference showcased the ideas of some big names in the Mars business. Amongst the most inspirational proposals were those of space tourist Dennis Tito (Inspiration Mars Foundation), who plans to send the first humans on a Mars flyby in 2018, and Bas Lansdorp (Mars One) who is calling upon the world to help create the first Mars settlement by the early 2020s. Whilst the feasibility of some proposals was questionable, there was no doubting the will and enthusiasm of those looking to expand the empire of humans out to the Red Planet, and beyond. With so much going on I’ll attempt to filter the proceedings to give an imaging science perspective, beginning with plenary speakers who may benefit from some machine vision.
There was no shortage of expensive image acquisition equipment, as John Brophy from the Jet Propulsion Laboratory described the on-going Dawn Mission to the largest of the asteroids. Whilst his presentation focused on the use of solar electric propulsion, he presented several stunningly hi-resolution images taken from the Dawn spacecraft of the surface of Vesta. These were contrasted with those provided by the Hubble Space Telescope, which were small and extremely blurry in comparison. The study of the thousands of Vesta images will occupy researchers as they search for evidence of the origins of the solar system on these ancient bodies. On the other end of the asteroid scale, Chris Voorhees of Planetary Resources described his company’s ARKYD space telescope, a partially Crowdfunded venture, which will soon be scanning the heavens for moving dots of light in their search for small resource-rich rocks which they hope to mine using robotic crafts, perhaps within a decade. In both cases, an automated assistant would be of great value in analysing the returned data.
On to Mars, Dr Nick Schneider from University of Colorado detailed the workings of the soon to be launched MAVEN mission, targeting the Red Planet’s upper atmosphere, ionosphere and its interactions with the solar wind. The orbiter consists of an instrument suite for various forms of spectroscopy, including an Neutral Gas and Ion Mass Spectrometer, and an Imaging Ultraviolet Spectrometer. These will be used to image the Martian atmosphere with reference to its chemical composition. Whilst down on the surface of the planet, Dr Jim Bell of The Planetary Society presented extremely hi-resolution images taken from the surface rovers Opportunity and Curiosity. The panoramic views around Gale crater from Curiosity could be zoomed in to incredible levels of detail, almost revealing individual grains, whilst the rover’s hand-lens could zoom in even further on selected samples. The list of high profile imaging devices and their endless streams of data presents a huge logistical challenge for those relatively few experts responsible for their analysis and interpretation.
Imaging played a significant role in some track sessions also. Jim Secosky, a former teacher, described his investigations into Martian ice deposits using HiRISE (High Resolution Imaging Science Experiment) data from the Mars Reconnaissance Orbiter. He used many views of the surface, some of which he personally requested as targets for the HiRISE team to image, in order to demonstrate changes in surfaces with seasons and slowly flowing dust buried glaciers. And then there was my own contribution, showing how surface area measurements of user-defined terrains could be extracted from HiRISE data in a quantitative way, including a detailed error theory for the prediction of measurement accuracies. With a generous 30 minutes granted to each speaker, I articulated my views and the views of my supervisor, Dr Neil Thacker, that making measurements using pattern recognition requires more than just off-the-shelf components and ROC style evaluations. With the aid of many slides and figures of histograms I explained how statistical methods, such as Likelihood and error propagation, could be applied to make quantity estimates with the associated error estimates which are required for useful scientific interpretation.
The non-vision related work was too vast to condense into this short report, but a great deal of the conference’s content can be found on the Mars Society website (www.marssociety.org), including some links to YouTube videos. The Mars Papers Archive will also soon be updated with full papers from all speakers. Details of our own work at the University of Manchester can be found at www.tina-vision.net and also on facebook/maptheplanetsproject.
I would like to take the opportunity to give a big thanks to the BMVA for providing a generous travel bursary, without which my visit to Colorado and chance to present to such a diverse audience would not be possible.