My personal research in the habitat is looking at optimization of resources and how to use our limited power and water most effectively to meet my crew’s needs. A lot of this work is intuitive. I note what we consume and learn what we need most, then compare our actions against what we have available. I use our current conditions and past actions to create predictions for the future. As inputs change (less resources are available) our actions have to change, too. Limiting negative impact to crew research and lifestyle is my goal.
A lot of good habits to save resources are the same as what you might do at home, like turning off lights when you leave a room, unplugging appliances that draw power in standby mode, and turning off the water faucet between uses. But a few things are different, like not doing laundry more than once a month and using no more than 8 minutes of shower time a week. We adapt. I use power meters to understand what appliances use the most power and try to manage our use on days with less power available. We don’t want to unplug our computers when we need to do work, and we have scientific equipment that needs to remain on even when power is low. We need to understand the habitat as an entire system, and cut out the unneeded resource suckers first.
In the past, engineering optimization programs have helped me find solutions in a more robust way. I’ve been working with the modeFrontier program to build a habitat systems model and use it to find optimized uses for our resources. My goal is to see how to use resources more intelligently and limit the negative impacts of resource management. The company that makes this software did an interview with me about the project and the HI-SEAS mission a few months ago:
So far, I’ve been able to build a simple habitat power model, but optimizing it has been difficult. Ranking appliance use based on personal preferences is a lot trickier than creating a lighter suspension component for a car. I’ve been experimenting with ways to quantify our interests and preferences, so I can tell the computer model how to rank and sort them, but adding people into a system always makes it a little more squishy. We aren’t so easy to simplify down to a ranked order of numbers.