It all started in Late 2013 when my wife and I started the SV-ROS Meetup. A Silicon Valley based Robot Operating System Meeting of “Birds of a feather” as a spin off of the Home Brew Robotics club. We meet the week before the HBRC last Wednesday of the moth.
The first SV-ROS meeting was at the Hacker Dojo, Wednesday November 20th 2013, and was very well attended about 34 people came and checked it out. Lots of interest and even a few of the OSRF folks showed up. This was amazing a very specific topic had so much interest and people in the area to attend. If we tried the meeting in Orange County where I was working with the RoboSub vehicle I would have had maybe 5 if we were lucky.
We had a few training session on how to do ROS on BeagleBoneBlacks, The Raspberry Pis did not have enough horsepower at that time to run ROS. But to keep people interested we needed a “Project”. Mardi found the Microsoft Kinect challenge which was Sept 15th 2014 in Chicago. The group was verify interested. So we spent the summer hacking away and we won First Place and a $5,000 prize. I did not even remember there was award money. That night in our hotel room, we wondered what to do next. Mardi again, found the NASA Centennial Challenge. Sample return robot Challenge SRRC. Which in hindsight turned out to be a lot of work and money. Could not have done it without the support of SV-ROS and HBRobotic Club members.
Matt Labbé’s of RtabMap write up of the event and comments about SV-ROS, https://github.com/introlab/rtabmap/wiki/IROS-2014-Kinect-Challenge
Was all about getting a robot together for a the SRRC’s June 7th competition. We worked out of the garage on Helweh Court, San Jose and test runs at a local park and at the Rose Garden. We built the robot by committee. Each person adding their experience to solving the problem in their domain. The day of the competition was the most stressful thing I had ever done and it was an absolute failure. We did not even get off the platform.
Fortunately, we had good company in the same situation and the teams that looked like they would collect all the prize money also did terrible. With a $1.5 million purse only $50,000 was awarded, the NASA and WPI folks decided to have one last run in 2016.
This SRRC attempt, with some hindsight, and a little experience. I focused on simplicity. We created Max with far less moving pieces. I created a simple gripper based on a corral being lowered over the sample and it closed onto a tray that would then lift and the sample carried to the platform. All tracking of the robot was done via a PT camera system located at the base that was watching the fiducial on Max telling him where he was. Max smarts was completely done by Girts Linde, He was truly dedicate and we lived together every lunch and dinner for 4 months.
Good enough to get second place because we got two samples on to the platform. and if WPI had put the pre-cache where they said they where going to put it we might have been in the money. Has taken more than a year, not to be pissed off about them being cheap. We did something really impressive, beating 44 other teams with hard work and zero luck. Looking back it was a great experience and I miss the all the people that gathered for the event.
We moved to Gig Harbor WA in June and pretty much was not thinking about robots and still trying not to be mad at WPI. We were so close. Really saw no future for MAX and was shifting over to Neato Botvacs for home inspection, object detection using spherical images. We then moved in November and set up house finally in home we owned and do not plan to ever move again. The new house has a great Studio. But still had so many distractions and not much ROS or robots.
Figured Max would be great at power washing the concrete. Want to do something to map the property we have 4.3 acres and about 1.5 acres of it is garden. Just knowing where each type of plant are what they are doing in time “seasons” during the year would be a cool project. Figured a drone and some VR images.
Once settled into the new house, new job, etc we had 7 sets of visitors over the summer and finally Paul Ekas invited me to participate in a JD industry picking competition in China that got me moving again about robots. I made a mini Max but being in Gig Harbor was a draw back. They are off to China hopefully to win something and gather invaluable experiences under the pressure of deadlines.
I have decided to try to work with some of the people at work and take on the Darpa SubT challenge. Have been gearing up for how to tackle those . Doing the planning process of laying out technology and how to be better than anyone else at the challenge. So pretty much trying to be fully engaged in making this work.
Is a few weeks away, Looking to have a Tracked Max, flying fiducial sand a cool mapping system in the dark, ready for the August Darpa Tunnel event. So much too do so little time…
Created a Mini Max version of SRRC max, Was to help with a SV team. but so far only the hardware and none of the software.
What can I say the Covid, year to forget? Not much was achieved in the way of robots in the PNW. Tried to generate some energy to compete in the Virtual Cave Challenge for the Subt but was not able to execute. We added a repo to test but far to complicated to get the bot reporting scoring. We did far more talking than doing.
Added a Nvidia RTS 2080 to my old Mac Pro and have it running ubuntu 20.04 with the newest Nvidia. Good way to repurpose a Mac that can no longer be software upgraded. The card is liquid cooled so added the pump and radiator to the external for the 2007 Mac Pro.
Did a lot with drones and to get working but so far lots of parts and drone but zero flying time. Have mini drones, Have a DJI drone, Have a 5″ FPV drone and lots of parts to make more but did zero flying time with any of them.
Started playing openMV cameras. but do not have a application to use.
Created a framework for a cable laying and pick up trolley for a robot. lots of pieces and need to focus on finishing it.
Probably the overall theme is to finish something, Lots of starts and parts but nothing to show for it.
What to get done this years? Will be a shift to use ROS2, drones, community and perception.