This ambitious rover project is a childhood dream of mine of building a small Sojourner like rover, fitting it in a rocket, and launching it out into a field where it will survive a couple of nights, or weeks, or however long it'll last. This blog is to document all the lessons learned, occasional failures, and all the fun along the way. To see how this all got started, go to my first post in May 2013.
Ok, so I haven't really done much with Inspiration in the past month and a half. I've kind of been taking a break on building and focusing more on rethinking ideas. I did manage to get some tests done before the sun went away for the next 5 months. (Well it doesn't totally go away. We have 8 hours of light but it's not very useful for what I need it for. Plus it's really darn cold out most of the time.) But anyway, I had noticed during our demonstration at SpaceFest that the inside of the rover got really warm when the sun came through a sky light and lit up the solar panel. So eventually I got around to testing the solar panel in direct sunlight and recorded a temperature of 115 degrees Fahrenheit on the under side. So I got to thinking, what is light weight and insulates well? I then looked down at the Styrofoam cup in my hand that was, very effectively, holding back the temperature of my morning coffee. Bingo! So after work I raced home and cut apart a Styrofoam container of some fantastic Thai food we had leftover from the night before. I cut some holes for the wires to fit through and stuck it to the underside of the solar panel.
I put the solar panel back on the rover and pointed it at the sun for a half an hour.
The results were a very comfortable 80 degrees Fahrenheit on all of the components inside.
About 3 weeks ago, I decided to get Inspiration outside and prove that the new suspension can do what I designed it to do. This was my last chance before winter so I apologize if the drive doesn't seem very well thought out. It was kind of a last minute idea to hurry up and record a video before the sun went down. But here it is, proof that Inspiration does, in fact, go outside and that the new suspension design actually works...except for a part I forgot to screw in place.
After we made the quick video demonstration, I started focusing on simplifying the code and making it more user friendly for the eventual graphical interface. Nothing fancy but a much needed command to stop before executing any other drive commands. The problem I had was that if you told it to turn and then mid turn, you told it to go forward, it would attempt to drive forward with the 4 corner wheels not pointed straight yet. So now the wheels stop at the beginning of each drive command and then it turns the wheels to the proper direction before proceeding with the drive. This should make it less prone to driver error and possible damage.
In the past couple of weeks I managed to join 2 different groups that will allow me to learn skills I will need for this rover project.
One is the Autonomous Snow Plow competition at the St. Paul Winter Carnival. I am on the planning committee and greatly look forward to seeing how people accomplish building a robot that can see and navigate a course on its own.
The second group I joined is a team of University of Minnesota students called LPRD who are designing and manufacturing a liquid fueled rocket. As of this evening, I have joined their launch sub-team and will be helping with fuel delivery systems for the engine. This is a very exciting project and they're always looking for help so if you're interested and can spend time working on rockets, click on the link I shared and sign up to help.
I hope to get moving on the rover project soon. SpaceFest VII is coming up in just over 6 months and I wanted to have some kind of lander put together by then. But first I need to finish the upgrades to the camera and metal detector. I think this is going to be a very busy 6 months.
For the past 3 weeks I have been scrambling to get Inspiration operational again in time to present at the Girls in Aviation Day at the St. Paul airport last weekend (9/26/2015). Working 9 hour days means I usually only get about 4 hours to work on the rover each night. It's not entirely flight ready, but the bulk of it is done, now, and that is a very exciting feeling! I can now focus more on code and sensors rather than design-test-redesign.
To catch up with the past month or so, let's begin somewhere after my last blog post. I think the last post explained how much better Shapeways was at 3D printing than our Zeepro 3D printer. So we ordered the rest of the suspension parts from Shapeways and waited; it takes about 2 weeks from ordering parts to get them in. Unfortunately, I made a mistake several months ago while saving a servo bracket design and ended up sending Shapeways a file that was no longer useful. So when all of the parts came in, I had 2 parts that didn't work with the steering servos. A rush to order the correct part ensued since the weekend of the Girls in Aviation Day was coming up in just a few weeks. While I waited, I used the 2 bad parts as a stand in so I could work on the rest of the suspension.
The new suspension required a longer shoulder mounting shaft to be cut to length and grooves to be machined in for the locking 'C' clips. I spent a few days going to different hardware and auto parts stores looking for the type of aluminum pipe from which I had made the previous shoulder mounting shaft, but didn't have any luck finding it. 5/16'' aluminum tubing is apparently hard to find. On a hunch, I stopped in at Ax-Man Surplus to see what they might have for aluminum tubing. After about 20 minutes of scrutinizing everything that might work, I found fiberglass tent polls. I asked for calipers to measure one and sure enough, an exact match! So now the shaft is cool-looking black fiberglass.
After mounting the servos and servo brackets to the shoulder joint, I had to put the body on. I then piled in batteries and electronics to see if the suspension could hold the weight... it couldn't. The spring inside of the shoulder wasn't strong enough anymore to hold the new electronics I wanted to add. So Cristin and I sat down one day to bounce ideas back and forth and came up with the idea of putting a "locking" latch on each shoulder. Now, with the latches, when the rover is in flight mode, the shoulder arms will be spread out. After landing, we can command the front wheels to drive backwards which will raise the body and bring the shoulder arms together where the latch would then catch and keep the arms from spreading apart again. This latch can be seen in this picture:
Some days during my lunch break, I squeeze in a couple of pages of Sojourner: An Insider's View of the Mars Pathfinder Mission by Andrew Mishkin. (yes, I've been working on reading this for a couple of years now during down-time. You can actually see a picture of it in a much earlier post.) Oddly enough, in my own way, I've been on a parallel path with Inspiration and Mr. Mishkin's book. I come across an issue with the rover and then read about a very similar issue on Sojourner. This was the case with the shoulder latch. I thought for sure this problem was unique to Inspiration because of the inexpensive and easy way I designed the shoulder. So to my surprise, after coming up with a latching idea, I read about a very similar latch that they had to put on Sojourner. So I must be doing something right if I'm unwittingly following in JPL engineers' footsteps.
Here is the suspension waiting for the electronics to drive it and the correct parts to come in the mail.
After getting the suspension together, it was time to give Inspiration a whole new electrical system.
Over the winter, I programmed a way for the Arduino to sense battery voltage for 2 lipo batteries and switch between the two as one would drain and the other would charge from the solar panel. Making these changes between 3 power sources required 3 relays and a 4th for main power on and off. It also needed 2 more relays to turn the power to the camera and metal detector on and off. To feed these relays, I needed a board for the batteries to plug into and a couple of power rails on that board to supply power to the different systems.
This new power distribution panel was also designed to eliminate the jumble of wires piled inside the rover's body which made access to the battery and components almost impossible.
Here's an older photo kind of showing this pile of wires:
Here is the new more organized wiring. Battery removal is MUCH easier, now, and access to the Arduino's USB port is possible if I need to upload a new sketch.
About a week before our presentation, Cristin came home and found me outside with this new power panel, 2 lipo batteries, and safety glasses on. She asked me why I was testing the power panel outside and I said that I didn't want the batteries to explode in the house in case I screwed up on the wiring. After some thought on how volatile lipo batteries can be, we decided that it was probably in everyone's best interest to not have the rover go up in flames during our presentation. (This was a mistake that could have ended our dreams of doing anything space related... or anything at all, really... had a battery exploded while pushing temperature limits)
We decided to fix this before it became a problem. I did some research on batteries and found these 7.4 volt 2600mAh Lithium Ion batteries and a nice little charger for them. They have a maximum temperature rating of around 450 degrees Fahrenheit which is way better than the 114 degree limit of the lithium polymers. These also have twice the power capacity so drive times are now around 2 hours per battery. The down side is that they are a little bigger and heavier... and needed to be in the rover within the next 4 days. We checked many websites and finally found one that had 2-day shipping on these batteries.
While we waited for batteries and parts, I started the programming for all 9 suspension servos. Getting the angles of the steerable wheels proved be a bit of trial and error. The Maestro controller didn't really work with degrees. Instead it worked with numbers from 5000 to 7000. 6000 being center but my brackets weren't mounted exactly straight so each corner servo had to have a different number for straight. One would be 6015 and another would be 5998. And then turning them inwards at around a 40 degree angle took even more tweaking. After a few hours, I finally got the lines of code correct.
Here is a video of the first time the electronics and suspension were integrated together and tested with the correct code.
Once I got the suspension operational, it was time to install the XBEE wireless receiver, replace the desktop power supply with a battery, and stuff it all into the rover body. Once I got everything back together and driving around on the floor, I couldn't help but set up a ramp to see how it climbs.
Here, I found that the slow speed setting was not providing enough power to the servos to get the front wheels to climb over the front of the ramp. Going at it in reverse worked out, though.
And then I felt daring:
Inspiration was now a capable rover. By Thursday of that week, the parts and batteries came in and I could finally take a breath. With Friday available for extra work, I decided to get the camera arm and metal detector arm programmed and operational. It's always impressive to see the camera rotate up from its stowed position over the solar panel and then pan back and forth.
Saturday came and we arrived at Holman Field in St. Paul at 8am. We set up our display complete with red sand, rocks, and information about Sojourner and Donna Shirley who managed the Pathfinder project for JPL. We had fun talking about how Sojourner *inspired* our little rover, and that's why we named it Inspiration.
Inspiration was a hit at the Girls in Aviation Day event. Kids and adults would hang out with us for long periods of time asking questions about Inspiration and engaging in some great conversation about NASA and all of the other space programs around the world that are currently working on robotic missions to the Moon, Mars, and asteroids.
So now that we have some down time, I can focus on getting other systems working. Next up for Inspiration will be to get the 2-battery system fully functional, design and build a new metal detector that will reach the ground with this new higher suspension, improve the camera arm, get the ultra sonic range finder on the front working, and then work on the GPS autonomy and graphical interface.
Lots to do yet so stay tuned and remember to chase the challenge!
Wow! Has it really been 2 months since my last post?! Well I've been up to stuff. I really need to be up to stuff now because there needs to be a working demonstration for an event next month.
So I've been busy the past week trying to make some progress since basically gutting the rover a month ago to rewire it . In the process of rewiring, I realized the new steering controller needed to be wired so I started figuring out how to do that. While I was figuring that out, I started printing the rest of the suspension parts so the things that I was wiring had a place to go. I got about half of the parts made and started having printer issues. The parts all had some kind of flaw that had to be worked around or ignored. I didn't like that so I decided to outsource all of the parts to www.shapeways.com after doing a sample run of their services.
You can see in the photo on the left, my misshapen part compared to the part I got from Shapeways and then the photo on the right is of the 2 parts I ordered. So needless to say, after I saw the quality of their parts, I immediately ordered the rest of the suspension parts. They should be here in a couple of weeks. So while I'm waiting for that, I started working with the parts that I do have.
I plugged in 5 servos to the Pololu servo controller and then plugged in my 7.6v LiPo battery. The fact that my servos are only designed for 6 volts at most suddenly dawned on me as one of the more expensive metal geared high torque ones started to smoke. Not a pleasant little puff of smoke but a rolling black cloud that smelled a bit like burning tires. Just a few seconds was enough to leave a nasty smell in the shop for a day or so. Here is a photo of the damage.
So with that servo now liberated of magic smoke, I grabbed my trusty power supply and set it for a more friendly 5 volts (something I should have done to begin with) and plugged it in to the Pololu servo controller. I then worked out some quick code for manually operating the part of the drive system that I had left. Everything seemed to work. Then we went to the hobby store to get more metal geared high torque servos and was informed that these might not be powerful enough for what we want them to do. Since I hadn't actually tried them with any weight on them, I really couldn't argue so we went home empty handed and set up a test of the steering with about a half a pound of batteries on it. Since last we weighed the rover, it was about 3 pounds, we divided 3 by 6 and got .5 pounds for each wheel.
Here is that test:
It seems like it should work. Carpet might be another story but with only a month before a demonstration, we decided to go back and buy the rest of the servos that we needed to finish the new suspension. Now we wait for the Shapeways parts to come in.
In the mean time, I need to finish the code for the rest of the servos, fix that wobbly wheel, and figure out how to fit everything back into the body. And also do some more research on these next couple of items.
This website gave me a great idea for another sensor on the rover. With this site and another camera with their infrared filter on it, the rover would be able to take pictures of plants and tell how well they are growing. Kind of beats just sampling rocks with a metal detector.
While searching for a way to have an arduino send back images from another camera, I came across this nifty little device used for multi-camera FPV (first person video) systems.
Since this little switch can handle 3 cameras, I figure I could have the main mast camera, the IR camera somewhere on the rover, and a hazard camera on the front. This could be a great little addition. I just need to find out if all 3 cameras will have power to them all of the time or not. If they do, I'll need to ditch the idea or work out a power switch for them.
And then lastly, this little 5 Volt 2.5 Amp voltage regulator that could replace the heavy relays that I currently have for switching between batteries.
These say they can be enabled or disabled with a high or low signal but it's not very clear whether that means the output can be turned on and off or just the regulating part can be turned on and off. I'll need to look more into this.
So that's about it for now. I'll try and post more now that I have a bit of a deadline. Some part of the rover should be operational again by the end of next month so that means I should have plenty to write about. So stay tuned and have fun out there!
Sorry for the month and a half long delay but... We're first time homeowners!
We've been busy painting
And doing yard work
We have a nice back yard that has filled in quite a bit since this picture. But I'd like to think there is a rover proving ground in there somewhere.
The shop is coming together. Just need to build a bunch of work benches and acquire a few more tools to make building robots a lot easier.
But since the last post, I haven't just been working on house things. I accepted a position as a Test Engineer at work so loads of training there and I also started work on the new steerable suspension. I spent a lot of time designing the new parts and even managed to print the left side components.
I finally figured out the right combination of settings for printing in ABS without a heated bed or enclosure. No more curling parts pulling off of the bed. A short 'raft' of PVA and then the ABS on top of that, along with proper nozzle height, seems to work great!
This is Inspiration with its new left side suspension. Still unwired and kind of mocked up here but this is the idea. A ton more ground clearance and better turning ability.
And just for kicks, I added a nice background from Inventor's stock background schemes.
That's it so far. I hope to start printing the right side parts this weekend and get it driving around soon. But in order to get it driving around, I need to change out the motor driver for the servo controller which also needs a little programming and also updating the Arduino's program so that it talks to the servo controller. After all of that, I plan on making a new metal detector arm since the current one is too short and will also hit the new front suspension. And then I'd like to change out the camera arm for something that can pan and tilt. Lots of things to do yet but it's coming together.
We also attended our first Tripoli Minnesota rocket launch day last weekend with my nephew Will and met some great people and got some great advice for the up coming rocket that will carry Inspiration out to its eventual landing site. Apparently there is a group of students that are attempting the same idea of launching a rover out into a field. We came late and missed them at the launch though. I'd love to meet them and compare ideas so hopefully they'll be around again.
Well.. stay tuned for more updates and have fun out there!
p.s.
If I could, I would have this SpaceFest timer running on the side of this blog. But for now, you get a link.
Can't wait! Time to SpaceFest VII
Alright I'm back. Slow progress. But that's been kind of the theme of this project since the start. Even though I'm finally done with school, it would appear that other things still fill my time. Currently there have been two other projects going on besides Inspiration. For one, we have been working on becoming homeowners! After looking at several houses, we've finally settled on one and now in the process of purchasing. If all goes well this month, Stray Robotics will be moving from this little work bench:
To this much larger work space with outlets already positioned for the super long work bench that will run the length of the wall and maybe that far wall as well.
The second project that's been taking up my evenings is a Shapeoko 2 cnc router that several of us Anoka Tech students are making. We got the kits, motors, power supplies, and hardware and started assembly a few weeks ago.
The frame as of last week
The stepper motors, motor driver, and power supply
Assembly as of 3/31/15
When I'm not working on those two projects, I try to get things done on Inspiration. Lately my focus has been on the front steering brackets and getting the Pololu Mini Maestro servo controller to do what the Arduino tells it to do. In this video, I have 5 of the 6 total drive servos plugged in and running along with 1 high torque steering servo. This is testing to see if the Maestro can drive all of these servos at once without any problems.
The Maestro performed flawlessly but there seems to be a delay on the Arduino side that I've noticed periodically while driving Inspiration around. For some reason this delay shows up more with the Maestro. I'm willing to bet it has something to do with my pieced together code and the way it looks for an input command from me.
But anyway, after this test, I started on the upper servo bracket. I drew up something to hold the steering servo onto the front suspension arm.
Once I had the bracket drawn up, it was time to send it to the printer. I didn't get video of this one printing but it turned out alright.
Here's a video showing how this bracket works with the wheel.
Just a mock up of how it will replace the current front suspension.
Next up for me will be to print the right front brackets and then try to figure out how to do the rear suspension. My plan is to have this thing steering by this summer, so hopefully I can spend some more time on this once the router is finished. Stay tuned and have fun!
Big things have been going on over here in the past month. Cristin and I (and no part of Stray Robotics) have been accepted into a fantastic and elite group of volunteers, The JPL Solar System Ambassadors. We, as individuals, can now go to schools and public events and discuss topics in Science, Technology, Engineering, and Math (STEM) or anything else JPL might like us to participate in. As an example, Cristin and another ambassador will be talking about Hubble at the Minneapolis Science Museum later in April. I have also taken a new job as an Engineer at a great company, so no more assembly or machine running type work for me. I get to finally put my Bachelor's and new electronics degree to good use! So a little inspirational tip for you timid folk like me, if you see a job you think you'd be good at but feel you don't have the qualifications for, go ahead and apply for it! You might be surprised by the people who interview you and find you have some great things in common. If you're a pleasant person and excited just to be there, they might be willing to train you. And chances are, you'll find people that think you'd be good at it too.
That brings me to the whole point of this project and this blog, which is to inspire everyone to "Go Forth and Be Brilliant", as +Cristin always tells me. What I try to do at the various places we show Inspiration is to inspire people, no matter how old they are, to pursue their dreams. Find that childhood dream that got lost during the course of growing up and go get it! It was your dream for a reason and I'm willing to bet that life will become much more interesting and rewarding once you start on that journey. I could go on about my life changing experiences since pursuing my childhood dream but I think that would be best shared elsewhere. And of course, feel free to share your stories in the comments or message me.
Now lets get back to the progress. We had a great meeting a few weeks ago with Donald (the LabView genius) and he's come up with a great way to retrieve specific information, that I can make constantly streaming from Inspiration, and turn it into wonderful gauges on the control panel. Gauges like battery level, internal temperature, servo positions, and anything else we want to monitor. He also figured out a way to grab the GPS data and use it in a Google Maps window to guide Inspiration to anywhere on the map. Very exciting stuff that is mostly waiting for the parts that I'm working on. Which is:
Steering. After noticing how terrible Inspiration turns on high traction surfaces, I've decided to make the corner wheels turn to make rotation a lot easier. This also makes Donald's GPS guidance system work a lot better since Inspiration won't have to be moving first in order to turn towards a way point. This also gives it about 4 more inches of ground clearance which was desperately needed since Inspiration couldn't even clear small rocks without getting caught on them. (The next Jenga block test with suspension was a miserable failure too and I probably didn't share that video... or even took the video.)
Speaking of videos! Here is a time lapse video of one of the steering brackets being printed on our 3D printer.
This was actually the second attempt because I'm still trying to figure out how to keep the material from peeling off of the bed during a print. The first attempt got about 1/3 of the way and then started to peel up so badly that it popped itself loose. I still have some techniques to learn yet before I print any more parts. But here is the finished bracket with the servos and wheel mounted on it:
I then decided that it was unnecessarily bulky and that it could lose some size and weight and add a little more ground clearance by taking off the useless tray that the drive servo was sitting on. I also shaved off some of the excess material in the middle of the bracket, making it look more streamlined. You can see the bracket changes I've made on this next picture (which, now that I look at it, I could have brought the ground up a little higher so it doesn't look so funny.)
I was also forced to bring the shoulder joints out farther because the drive servos stick out a bit and would hit the body during any kind of suspension movement. So Inspiration now has a wider stance too which could probably come in handy. But it also means that I'll need to make the rocket a little bigger too. So far it looks like the rocket might be about 10 to 12 inches in diameter and that would make it roughly 8 feet tall if I try to keep it in proportion. But that's a much later problem to deal with.
Right now, the focus is on steering and a new suspension that goes with the new brackets. Then the best part, test driving! So stay tuned and share some stories or ask some questions in the comments. And remember to Go Forth and Be Brilliant!
