Camp Outback, TX – The drive to finish the min-challenge caused the team to push CajunBot a little too hard. In the process we ended up wounding her.Sunday morning we were all upbeat that CB was ready to take on the forty miles run. The previous day, Adrian had said he would survey the remaining course if CB finished the 3.2 miles course. It turned out that he had to rush to complete the surveying for CB sort of snuck up on him.
While I was happy with CBs performance on Saturday, Pablo and Suresh were not quite so. They did not like that CB was not able to maintain a graceful gait while going over broken rocks at high speed. They had asked Joshua and Santhosh to get PID values for steering CB at different speeds. This work was already completed on Saturday night. So Pablo and Suresh went on with modifying the system to use different sets of PID values for different speed. The results were visibly good. It was like CB had been transformed overnight. The jerky, robot-like, movements had gone away. She was making very smooth turns.
Excited with all the progress, we ventured on to run the eastern track surveyed by Adrian. The travel started pretty smooth. Even though the track was barely eight feet wide, CB was cruising along very well. Then at some point we found that CB veered away from the track and started running parallel to the track, going through grass and bushes. Joshua took manual control, got her back on track, and let go. On she went back into the bushes.
This behavior got quite puzzling. We were already about two miles into the track. It did not make sense to drive the bot back manually. We figured we may as well run the bot on her own and bring her back. The track was a loop, anyway.
After making several close calls in the rather narrow track, we finally reached the ‘main’ road of the ranch. This was a little more wide, about 10ft. Having driven through a much more cramped space, we were finally relaxed. We were barely half a mile from the camp.
Just what happens when you get numbed by a continuous error. We had begun accepting CBs running parallel to the track as a normal behavior. There was another close call, when CB almost ran into a big rock hidden behind grass. As always Joshua was quick on the RC, taking control of CB before she kissed the rock. He brought her back on the track and let her go.
By now it appeared CB was getting tired of being controlled by RC. She started running on the track, instead of in the grass. We were happy with this welcome change.
Then all of a sudden, I see CB slamming straight into a tree. No warning, nothing. A fraction of a second ago she was going smooth, orienting herself to drive past the tree. But for some inexplicable reason, she decided to turn a little harder and run straight into the tree.
While Joshua was still collecting his nerves, Pablo and Amit were on their fours, looking under CB’s belly, surveying the damage.
Diagnosis – CB had broken an axle.
Once we had our senses back, someone observed that the computers and electronics were still humming. The severe shock had not blipped them. In the past even a slight bump would trip our computers. This is thanks to the new structure designed by Joshua H and Chris L, manufactured by BEGNAUD, and the double redundant shocks mounting offered by Hardigg Case.
It was suddenly time for everyone to pack and go back home. There is not a whole lot one could do with the broken axle in the ranch.
A test is said to be successful if it makes a system fail. With that measure our field test was resounding success. Thought we did not complete the forty miles mini-challenge, we were quite happy with the progress that we made in the short five days.
This message is brought to you courtesy of satellite internet connection provided by SOLA Comm.
– Arun Lakhotia