The year is 1994. NASA and Carnegie Mellon University have spent nearly $2 million to build "
Dante II," an eight-legged walking machine designed to enter the crater of the Mt. Spurr volcano in Alaska. We were staging the robot at the University of Alaska-Anchorage, and it was the day before final deployment to the base of the volcano. During the last reconnaissance trip to the mountain, we noticed two items as we flew over the crater: there was a lot of infall into the crater from the upper slopes of the mountain (the active crater was a secondary vent on the side of the mountain) which created a lot of mist and fog inside the crater as the material was melted by the escaping volcano gases, and we thought we saw some reflections from the bottom of the crater, leading to the idea that there might be a pool of water at the bottom of the crater.
The laser and its spinning mirror scanning assembly were housed in the purple cylinders at the top of the robot. The problem is that all the mist and fog would collect on the lenses of the laser scanner being used to map the terrain around the robot (the terrain maps were needed so the robot could figure out where to place it’s feet as it walked down the crater walls), and make them unusable. We needed something to protect the scanner from the falling material and floating mist. And we needed it quickly - we had about four hours before we had to start loading the robot on to the helicopter for the flight out to the volcano. What to do?
Time to run off to the Eagle Hardware store in downtown Anchorage. Grab a 30-gallon plastic trashcan with lid, and run back to the lab as quickly as possible. Throw away the trashcan (which makes for its own unique problems – have you ever tried to throw away a trashcan?), and keep the lid. Cut a round hole in the lid just big enough to fit over the scanner housing, and hot glue it on top of the robot. And that explains the big floppy black rain hat-looking thing at the top of the sensor mast on the robot. It worked great, and kept the laser and scanning mirror clear for the entire time the robot was inside the volcano.
On to the other problem. If there were a pool of water at the bottom of the crater, then it would be important to try to collect some of it for later analysis. The water would have dissolved volcanic gases in it that could lead to a better understanding of what was happening deep inside the volcano. This was very important to the science team. Unfortunately, the robot was never designed to collect water samples and we had no mechanism for collecting one. Just like the rain hat, if we were going to make something it would have to be done quickly.
Enter an empty Evian water bottle, the pull cord from a lamp, a piece of a vacuum cleaner attachment, a round wooden ball from Eagle Hardware, a plumbing fitting, and more hot glue. Voila! A liquid sample canister with a one-way valve that could be dipped into a puddle and would retain any collected water without spillage. Duct-tape it to the front left leg of the robot (look very closely at the picture and you can just see where it is taped in place), and away we go.
Both of these last minute additions worked perfectly. And it proved once again that “Apollo 13 engineering” techniques can work wonders – finding ways to use what you have on hand to solve a problem, by getting creative and discovering uses for items that were never intended (sounds like the early days of FIRST).
Of course, that was of little consolation when the grizzly bear gnawed on the robot the very next day, but that is another story…
-dave