User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

The Great Mojave Desert Rover

Sat Feb 24, 2018 2:14 am

Hello, first-time poster long time Pi'er

Update: I have created a blog for this project: http://rover.bimthoughts.com/

The Great Mojave Rover Project is to build an autonomous rover that I can leave in the middle of the Mojave Desert and explore remotely.

Design Considerations:
* Utilizing existing JPL rover design for guidance
* Designed to be remotely controlled over the internet using a "go there" approach.
* 64kbps Cellular internet connection (T Mobile IOT deal)
* Mesh network if a desire to explore a canyon
* GPS, Temperature, Humidity, Compass, and gyro recording
* Photogrammetry to map terrain
* Tip alarm
* Solar battery for charging NiMH batteries
* 3D printing
* Weatherproof
* Six wheel drive
* Slow and steady
* Leave in the desert for a few weeks at a time to explore.

The stage I am in now is design and prototyping the hardware. I still have a bunch of printing and design to do before I am ready for elec work.

Right now the size is 375mm x 250mm. Trying to keep is small.
2018-02-23_RoverDesign.jpg
2018-02-23_RoverDesign.jpg (128.23 KiB) Viewed 3778 times
Last edited by BIMThoughts on Sat Mar 10, 2018 6:25 pm, edited 4 times in total.
Bill
Learn more about the rover: http://rover.bimthoughts.com/

IanS
Posts: 246
Joined: Wed Jun 20, 2012 2:51 pm
Location: Southampton, England

Re: The Great Mojave Desert Rover

Mon Feb 26, 2018 4:35 pm

Depending on your balance between using time and existing facilities (e.g. if you already have 3d printing capabilities) and spending money, you might want to look at www.turtlerover.com. It is a four wheel rocker rather than six wheel rocker-bogie platform, and you will still need to integrate some of your requirements such as cellular and solar, but it looks like it would give you a serious head-start.

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Mon Feb 26, 2018 5:37 pm

I did look at the Turtle Rover and signed up for the open source part of the project. It looks great. But I have the printer and the time to play, I guess that is part of the fun.

I think of this project of an experiment of can I do this. So far the hardware and the mechanical side is coming along great. I guess that is the easy part for me. The hard part is still around the corner, Solar and Photogrammetry.

I look forward to posting more about this project and hope to get some people interested. It helps to have people interested to keep me going.
Last edited by BIMThoughts on Sat Mar 10, 2018 4:27 pm, edited 1 time in total.
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Sun Mar 04, 2018 5:08 pm

Progress Report.

I have started printing out parts. Why in pink, well I have a bunch of the stuff when I purchased my printer.
IMG_0527-1.jpg
IMG_0527-1.jpg (64.31 KiB) Viewed 3778 times
IMG_0528-2.jpg
IMG_0528-2.jpg (81.05 KiB) Viewed 3778 times
Last edited by BIMThoughts on Sat Mar 10, 2018 4:27 pm, edited 3 times in total.
Bill
Learn more about the rover: http://rover.bimthoughts.com/

IanS
Posts: 246
Joined: Wed Jun 20, 2012 2:51 pm
Location: Southampton, England

Re: The Great Mojave Desert Rover

Sun Mar 04, 2018 8:46 pm

Why in pink? Because when the battery goes flat and you only know its location to 100m, you want some chance of finding it!

User avatar
Defiant
Posts: 179
Joined: Tue Oct 30, 2012 6:17 pm
Location: Hamburg, Germany

Re: The Great Mojave Desert Rover

Tue Mar 06, 2018 12:08 pm

Sounds cool. Having done a few robots myself I would like to point out two things:

- The motor in your picture "looks" (I don't have it specs) a bit undersized for a 40x25cm robot. Note that it is hard to change the engines later because it influences other parameters like space and battery lifetime.
- Determine your wanted battery lifetime first so you can calculate the needed battery size. On a robot you are limited on weight and size. At the end of your project you can't just double your lifetime by doubling the number of battery cells because the motors won't be able to lift twice the load..

Also charging from 40x25cm solar panels probably won't make you happy.

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Tue Mar 06, 2018 4:36 pm

Yep figuring out power management and time to charge is all part of the learning curve. I am thinking of having about 7-9 watts of solar, so my plans for the size of the body have changed a bit. The body is a bit longer now. My printer (MOD-t) will only print 100 mm x 150mm x 100mm so I have created sections of the body that are 90mm x 140mm. This will allow me to grow the body as needed.

One of the criteria for this project was to set destination and let it go, then wait for charge or orders. I don't plan on driving this thing more than few times a day. I would like it to spend time photographing and such. hopefully I can get a 3D map of the area from it. As far as what else will it do, I don't know that yet.

I should have some solar panels in today to test. I am going to see if I can power the Pi from one of them. I will test by pinging the PI all day. After that time to figure out batteries and charging and seeing if I can keep it alive all day and night. During those test, I am still printing and designing all the parts. I am also thinking of other things that can help save battery. I am thinking of using a relay to turn on and off the servos, turn on the relay, send power to servos/motors and turn it off when movement is done. Idle electronics taking power for no good reason is a bad thing.

Oh, and I need to test how well PLA will hold up outside. I have read multiple stories on the Internets that contradict themselves, I know right.
There may be an ABS printer in my future.

Lots of time ahead of me.

Bill
Last edited by BIMThoughts on Sat Mar 10, 2018 4:28 pm, edited 1 time in total.
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Tue Mar 06, 2018 7:20 pm

Solar update.

I was able to get the Pi to boot and run with (2) of these panels. http://amzn.to/2HbVecX they are 2.5W 5V/500mAh.
I was also able to keep it running with only one panel in full sun.
NOTE: this is an experiment, not a suggestion to get the panels above.

I tested with a WiFi ssh connection.
IMG_0541.JPG
IMG_0541.JPG (142.84 KiB) Viewed 3778 times
Last edited by BIMThoughts on Sat Mar 10, 2018 4:29 pm, edited 1 time in total.
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Tue Mar 06, 2018 10:39 pm

Counter-rotating Differential mock-up.

This is the gearing that will keep the rover body "level" to the ground.
It is mounted in a first draft print of a body section. Pi Zero in there for size.
I have not decided on a Pi Zero or 3. I am starting to think that the Zero will be a better bet. Don't need the processing yet, but do need the power savings.


Very happy with how well the gears meshed the first try.
IMG_0544.JPG
IMG_0544.JPG (84.86 KiB) Viewed 3774 times
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Sat Mar 10, 2018 4:36 pm

Yesterday I started a real blog at Http://rover.bimthoughts.com where I will add more information including code and such.

I also tested run time on 4 AAA NiMH batteries, I was able to get 21 hours of runtime. This is not too bad for cheap Ikea batteries, and if the solar can keep them charged I may be in good shape.
Full write up on the battery test here with Python code: http://rover.bimthoughts.com/?p=39

side note: I am going to keep the compute and propulsion systems separated, using different solar and batteries for each system. I will also isolate the different compute systems as well.
BIMThoughtsAAABatteryTest.jpg
BIMThoughtsAAABatteryTest.jpg (111.48 KiB) Viewed 3640 times
Weird note: I noticed that the labels on the batteries in the package are all the same direction and rotation.
Last edited by BIMThoughts on Sat Mar 10, 2018 4:52 pm, edited 1 time in total.
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Sat Mar 10, 2018 4:51 pm

I was finally able to get a mock-up of the rover suspension today. I am using Pink PLA because I have an unused 1kg spool. For the final print, I will most likely use white. Need to reprint the body section and wait for 2 more tires to arrive from China. before I can assemble for roll test 1.
RoverSuspension.jpg
RoverSuspension.jpg (101.24 KiB) Viewed 3631 times
Notes:
The final version will be connected via 8mm carbon fiber tubes with the wires running inside the top tube into the body.
Raspberry Pi Zero W in the photo for size

more information here: http://rover.bimthoughts.com/?p=64
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Sun Mar 11, 2018 1:26 am

More progress today. Lots of time waiting for the printer. Still missing two tires, but they will be here soon. Need to order 4 more for sandbox rover. I am waiting to receive the other order as I want to make sure they are the same tires and they are really not glued to the supplied wheels.
Learn more here: http://rover.bimthoughts.com/?p=85
IMG_0553.jpg
IMG_0553.jpg (145.84 KiB) Viewed 3577 times
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Sun Mar 11, 2018 11:33 pm

Spent the day redesigning after the test of the AA batteries and the Raspberry Pi Zero W. I am hoping that (2) of the solar panels that are producing up to 5 watts will be enough to charge the AA during the day. I made the rover body/pan modular, this allows me to add to remote sections as needed to make it longer.
I need to figure out where I am going to put photo arm, I am thinking middle but there may not be enough room with the Diff.

The other large variable is what batteries to use for the motors for the amount of traveling I want to do a day. I don't plan on driving this thing very long during the day. If I get 15 - 30 min of drive time I will be happy. I don't know until I get it rolling how much power it takes. The idea is to travel take photos, build 3D map and repeat.

Size is now 275m x 520mm x 190mm

Once I get it rolling, I will post the STL on Thingverse.
rover side 2.png
rover side 2.png (100.23 KiB) Viewed 3505 times
rover version 2.png
rover version 2.png (93.62 KiB) Viewed 3505 times
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Thu Mar 15, 2018 3:26 am

Progress today.

Got some more parts printed.
IMG_0568.JPG
IMG_0568.JPG (119.57 KiB) Viewed 3434 times
The suspension is working and the diff is keeping it level.

Need to work on the Servo Programming now and then get it rolling under it's own power.
Right now I am using wood sticks, but the real version will be printed with PEGT and use Carbon Fiber tubes.
Bill
Learn more about the rover: http://rover.bimthoughts.com/

pootle
Posts: 351
Joined: Wed Sep 04, 2013 10:20 am
Location: Staffordshire
Contact: Website

Re: The Great Mojave Desert Rover

Thu Mar 15, 2018 9:52 am

As long as you use the right gear ratio I would think the micrometal motors would be fine. The 298:1ratio are pretty torquey and there's always 1000:1. My big worry would be sand and dust.

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Thu Mar 15, 2018 1:19 pm

Right now there are (6) 200:1 motors. I too am worried about dust. I am going to add a dust shield around them.

After taking a look at it this morning. I may need to change the single tube idea to a two tube idea. There is way too much wobble.
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Fri Mar 16, 2018 2:19 pm

Today is the big solar test day.

Last night I hooked up two solar panels and 4 AA NiMh batteries up and powered the Raspberry Pi Zero W all night last night. This morning the Pi was still alive, if all goes well the batteries should charge enough to make it through another night. I am going to leave it out there until it dies.

This test is very unscientific if it works it's dumb luck.

Update: I was able to get 27 hours out of the PI Zero W. with batteries and solar. I did not make it through the second night. I am either going to need more solar or I did not wire it right. I am going to look at the wiring in the morning.
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Sun Mar 18, 2018 7:18 pm

Did the first test run today and found a design issue with the suspension. After a bit of more testing, I will post a video and my findings. But it drives, a lot faster than I thought. It also used a bunch less current that I thought it would. .7 amps at speed and 1.8 Amps to speed from a stop.

I also have a PI Zero still running for 21 hours on solar and 4 AA batteries after redesigning the solar charging circuit. fingers crossed.

I have video footage of the first drive on my site: http://rover.bimthoughts.com/?p=122
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Fri Mar 23, 2018 11:24 pm

I created a PHP, MariaDB, and Python tracking program this week. I will be fully documenting it on the rover site this weekend, from start to finish. It seems like overkill for tracking uptime but it's the foundation for sending and receiving data from the rover to home base.

Picked up a stepper motor yesterday as well, may give that a try for replacing the baby N20 motors. But I must say I am happy with them so far, they don't draw much power and they are fairly quick.

I am also still working out solar charging and powering of the Py. I was able to get over 32 hours, just need it to get through the second night.

Update: Finish a 3 part blog post on how to record uptime using Apache2, PHP, MariaDB, and python using 2 Raspberry Pis. Once Pi 3b+ as Mission Control the other a Pi Zero W as the RoverPi.
You can find that at http://rover.bimthoughts.com

direct links:
http://rover.bimthoughts.com/?p=149 INSTALLING APACHE, PHP AND MARIADB ON A RASPBERRY PI 3B+
http://rover.bimthoughts.com/?p=171 SETTING UP THE DATABASE
http://rover.bimthoughts.com/?p=189 RASPBERRY PI RECORDING UP TIME TO MISSION CONTROL
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Fri Mar 30, 2018 2:02 pm

Starting Solar test 4 today, Last time I was able to get 30 hours out of 4 AA NiMh batteries. Today I am shooting for making it through the second-night barrier.

This time I am using three 1N5817 Schottky diodes to keep the energy from going from the batteries to the solar when the solar is not producing enough power.

On an different note, I did see a great post that is of interest to this project as well, Raspberry Pi UPS with Auto Restart this may come in handy. I am also thinking of an emergency out of power call home, this would use the propulsion system's batteries as backup. Too many thoughts, I need to get the first goal objectives done first.

If you would like to find out more information about my project and it's progress head over to http://rover.bimthoughts.com
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Sun May 13, 2018 4:49 pm

Back with an update: Solar Test 6.
SolarTest6Camera.JPG
Solar Test 6 Unit
SolarTest6Camera.JPG (153.15 KiB) Viewed 2537 times
Took a bit of a work break, but I am back solar testing. I was able to get the Raspberry Pi to work for a week with four batteries. Now I am again trying to get it to work with only three batteries. I am using (3 each) 18650 Li-ion batteries, 03962a chargers and 5v 500 mAh solar panels. I am also using a Pimoroni LiPo Shim; this little device is great by the way. I decided to move from AA NiMh batteries to the 18650s because there are cheap and have chargers available, and I am hoping that these cells will be safe enough to survive the heat of the desert. I will be testing outside for most of the summer before launching the rover. I still have a bunch of work to do.
SolarTest6Wiring.jpg
Solar Test 6 Wiring
SolarTest6Wiring.jpg (110.58 KiB) Viewed 2537 times
As of this writing, I fully charged the batteries the placed the test unit outside last night. It made it through the night and so far this morning I have one blue light on the charger and two reds. I hope that means that I have one charged battery and the other two are charging. I also covered the batteries with white paper and tape; they were getting a bit hot in the last testing in direct sunlight.

I have few differences in this design over the previous.
  • I used a bus to wire the solar and batteries, see the photo above. I soldered the pins together vertically. It's unnecessary, but it makes the wiring look much cleaner.
  • I put the solar panels in parallel. The thinking behind this is it will allow more power to charge the batteries should one or more of the solar panels are in the shade. I am also hoping that after one battery receives its full charge more of the power will go to the others.
  • I put heaver wire on the battery holders, the existing wires only had 7 strands of wire in them, I replaced them with 24 gauge silicone coated wire. If you have not invested in silicone coated wire it's the way to go, you can strip them with your fingernail.
I would like to have two sets of 3 batteries. The first set for Compute the other for Propulsion once I get to that stage. I am also thinking of putting all six solar panels in parallel, but I fear that may be too much current for the chargers when in full sun.
SolarTest6.jpg
Solar Test 6 Schematic
SolarTest6.jpg (151.65 KiB) Viewed 2537 times
Please note: I am not an expert and use this at your own risk.
Bill
Learn more about the rover: http://rover.bimthoughts.com/

User avatar
BIMThoughts
Posts: 22
Joined: Sat Feb 24, 2018 2:02 am
Location: Las Vegas, NV
Contact: Website Twitter

Re: The Great Mojave Desert Rover

Fri Jun 29, 2018 1:56 pm

It's been awhile since my last update. I am in the middle of my 9th version of testing Solar. My previous record was four nights. Last night was the 5th glorious night for solar test #9.

After receiving inspiration from the "Solar-Powered Squirrel Kam" project. https://www.hackster.io/reichley/solar- ... ted-797db4 This time I connected an Arduino clone of the Pro Mini to the LiPo Shim to shut down the Raspberry Pi Zero W overnight and restart it at daylight using a light sensor. I am feeling very confident in this new setup. I am getting a full recharge of the (4) 18650 batteries in about 5 hours of from (3) 5v 2.5-watt solar panels.

Now that I know how many solar panels I will need and the batteries I plan on using, I can get back to designing the Rover. I figure I will need two banks of eight 18650 batteries. One bank for the compute system the other bank for the propulsion system. Each bank charged by three solar panels. As far as linking the two banks together. That might be a good idea, but I have not thought that far in advance yet.

Steps for this weekend are to replace the Pro Mini with an Adafruit Trinket M0. The M0 may be the way to go as I can program it in Python and copy the updated program to its disk. I won't need to worry about the Arduino IDE, or it's CUI utilities to program it.

Next week is a road trip across the United States, I am planning on using the free non-driving time to finalize version 1 of the rover chassis. Then start the designing version 2 of the rover. I hope to post the STL files for version 1 next week as well.
Bill
Learn more about the rover: http://rover.bimthoughts.com/

Return to “Automation, sensing and robotics”