Analog Experiments and Reality

As most readers will know by now, we are conducting analog space exploration research at MDRS, but not conducting formal "sims" that require wearing suits. Our interest this year is not EVAs, but what's happening inside the hab itself, and in particular providing tools we can use when inside a space ship or habitat to interact with automated systems.

This idea of "analog research" bears some thought and explanation.

The Mobile Agents project first came to MDRS in 2003 after experiencing dust storms near Winslow, AZ (Meteor Crater) with NASA's DesertRATS Project in September 2002. We were working under tarps, trying to debug our system in a realistic environment, which included a pressurized suit. This combination -- dust, wind & record rain, a 20 to 40 minute run time (for the suit), and an early prototype system -- was untenable. Someone from JSC said, "Not ready for prime time," which stung. It was unfair because the suit's operating overhead to useful runtime was about 10 to 1, and we never viewed our work in the desert as a demo, but as a test, with many experimental aspects that (as we discovered) could take days to set up. Trying to synch up with the suits was a losing proposition.

Having managed Crew 5 at MDRS in April that year, I realized the Mobile Agents team should regroup here. MDRS attracted me for two very important reasons.

First, the hab would provide a protected place for our computer systems, as well as a comfortable place to collaboratively debug and refine our designs. By living in our work area, we could put in especially long days. In that respect, the moniker "research station" appeared both prescient and inviting for our needs.

Second, I wanted an authentic site for exploration. Having been at MDRS for rotation #5 (including a reconnaissance with the Zubrins and Chris McKay in early March 2002), I had learned of ideal geological sites for testing the wireless system. Thus in April 2003 (Crew 16) and April 2004 (Crew 29) we ran trials of Mobile Agents around the habitat, in the area to the immediate south, and in Lith Canyon. Last year (Crew 38) we shifted to another area that was still new to our geologists, just to the east of the hab.

In these experiments (once we left sight of the hab), the geologists were actually exploring, using Mobile Agents to plan, document observations, and share data about a new place. The land around MDRS is of course called a "Mars analog" because of its appearance, isolation, and desert qualities. The experimental uses of Mobile Agents were "Mars analog EVAs" by virtue of occurring in a Mars analog setting, involving actual exploration, and operating under EVA- like constraints. These constraints included: Formally planning the EVA with a remote science team, following the plan, processing data (GPS, biosensors, photographs) in real time, and interacting with a robotic assistant with voice commanding. Also, to the extent safety allowed, we treated the EVA area like a movie set, with the support team out of sight on a 180 degree camera line.

To carry out this research, none of our experiments required the support crew to be wearing simulated space suits. Instead, the experiments consisted of what I believe to be the essential elements of analog research: 1) authentic work, 2)useful prototype systems (similar to what might be used on Mars EVAs), 3) controlled protocols, and 4) systematic documentation of what happened (e.g., every remark uttered by the geologists is recorded during the EVAs by the voice commanding system, which allowed me to create videos in which we can hear the agents and astronauts together). What aspects of the MDRS rotation need to be controlled (part of the simulated EVA) are a matter of judgement, just as for an experimental setting. In this respect, MDRS and the surrounding area constitute our laboratory for doing space exploration research.

Our work with robotic systems continues under different ESAS (Exploration Systems Architecture Study) programs, with a field demonstration planned later this year. Now in our last year of funding for the Intelligent Systems Program (which began in Fiscal Year 2002), and working also within an ESAS program devoted to "spacecraft and vehicle autonomy," we shifted our focus at MDRS to the power system.

One difference in our work this year is that instead of having two geologists using the system, everyone living and working in the hab becomes part of the experiment. As we have been working here today, with the generator out and the batteries dying, we have realized additional inquiries that need to be part of the system (e.g., "Tell me when the battery VDC drops to 22 volts").

As part of an analog experiment, we have had some paradoxical situations: We needed to reduce our power as much as possible, but we needed some computers to run to record our actions (Crew Activity Analyzer), to run Mobile Agents so we could inquire about the system status, and to continue developing improvements to the system. So we are literally living and working within an analog experiment.

Indeed, I need to be spending time with the engineer now, to document how he is debugging the replacement generator, which has been unable to synch with the generator.

Bill Clancey
Commander, MDRS Crew 49