The design for the EcoDisaster barrel lifter is complete and has been tested. We’re quite pleased with how it looks, but very unhappy at how it works.
The drive from the servo to the jaws (which is via fishing line) is way too vague. As a result the barrel slips out at the slightest provocation.
There’s going to have to be a plan B.
Don’t you just love Fusion 360!
Still Dad continues with the design. As you can see the barrel grab has been worked on. It uses two micro servos, one to open and close the jaws and the other to lift the barrel (by about 10mm). We’d love to cut the jaws from aluminium (because they’d look amazing), but I guess we’ll make do with 3D printed.
We’ve added a couple of sideways facing time of flight (ToF) sensors to help with the maze. We only used a front mounted ToF last year because we could correct any turn errors by targeting the alien faces with the camera, but camera feedback looks less viable with the new maze game. Dad’s a little nervous about this; last year the ToF caused a world of problems until he dropped an arduino between it and the RPi. He never really got to the bottom of why it was unreliable, but we do at least have a work-around if we need it.
We’ve also got a rather natty camera mount. We want to be able to mount the camera high above the chassis for Eco-Disaster and Minesweeper but right on the deck for all the other games. So we’ve 3D printed a camera mount that can be unclipped, then clipped back together with an extension piece inserted. This took three versions to get right; it’s not as easy as it sounds!
Oh, and yes, we’ve managed to loose the extra length in the design that was troubling us last week. Happy days!
You may recall (from our last blog) that our first objective is to make PiDrogen smaller. To that end Dad has been busy with Fusion360 designing P20; here’s a screen capture…
The good news is that the design is well under way and it is about 20mm narrower than P19.
P19 was a mixture of 3D printed parts and laser cut acrylic, but since we now have a 3D printer, we thought we’d try 3D printing the whole thing and the design takes that into account. This is a good opportunity to try this since there’s still time to start again if we have to. Once we have the chassis components printed we will install radio control then the boys can “test” the chassis. Hopefully it will last longer than the original P19 chassis which had a broken back axle within about 10 minutes of Nathan at the controls.
As I said, that’s the good news. But there’s also bad news; the new design is currently longer than P19 (by about 20mm). Oh well, back to the drawing board, sorry, CAD.
Following this year’s contest (in March 2019), we had a team debrief to decide what we could do to improve the performance of PiDrogen in the future. Here’s what we came up with:
- We think it would be good to make the robot smaller. The current robot (“P19”) is only slightly narrower than the rules permit (225mm) and Nathy feels this slowed him down on the obstacle course. However, we do not want to compromise the obstacle clearing capacity of P19 so we are going to keep P19’s big wheels.
- We also want to make the robot faster. Many robots were faster than ours at the competition, although, not all appeared easy to control. We plan to experiment with different motors and gear ratios to ensure that we can still exercise precise control with increased speed.
- Whilst looking at speed and control we will probably rework P19’s video control software. 2019’s autonomous competitions were all conducted at 25% motor power, so we’re hoping for some big performance increases.
- We didn’t do well in Pi-Noon this year; we were knocked out in round two. We are going to trial the use of mecanum wheels so that P20 (PiDrogen’s evolution for the 2020 competition) can move in any direction and rotate simultaneously. The hope is that the robot will be able to quickly move around the arena but always face it’s opponent. This is going to be hard to drive, but Nathan (team PiDrogen’s driver) can fly drones and we will configure P20’s controls like those of a drone. This requires a motor controller rework since we will need four independent motor control circuits.
- P19’s RPi and touch screen are mounted on it’s back. This was done to allow robot configuration and start/stop buttons to be accessible. This turned out not to work well as some widgets in Tkinter cannot be resized and are therefore are too small to use with sausage fingers on a 4”screen. This location makes the RPi vulnerable so P20 will have the RPi within it’s hull.
- Since the touch screen will be removed, configuration will take place via a WiFi connection with a wired ethernet backup. We think we might build a USB connected screen with a few buttons to provide start/stop/game select/status feedback type functionality.
- Finally, we have some plans to improve P19’s Nerf gun (which is too big and not accurate enough). We are going to try reducing the size of some of the components and mounting the laser sighting module with thumb screws to permit delicate adjustment. We are going to try reducing the length of the darts and possibly weighting them so that they fly truer. We are going to try casting new flywheels with built in spirals to spin the darts laterally as they fly, also to make them fly truer. Lots of experiments will be needed on this and we need to find a way to measure the accuracy of the gun in an objective way.
Should be ready Monday week then…
I know it’s a bit behind the curve, but WE’VE BEEN ACCEPTED INTO PIWARS 2020! Whoo-Hoo
I suppose we’d better get building then…
We took some pictures of PiDrogen before it went to school for Show’n’Tell. You know, just in case. So here they are.
Amazingly it came back unscathed! Yay!
I know this post is rather late but WE WON! Well, we won the Advanced Category that is. You could have knocked us over with a piece of heat shrink.
Here’s our very lovely trophy…
We want to say a massive thank you to everyone who made Pi Wars possible, especially Mike & Tim: we had a lovely time and we can imagine how much work it was.
Now, please can we do Pi Wars 2020? Perhaps the theme could be “perfect vision”?
PiDrogen has been completely designed using fusion 360. We created this render for the programme submission.
We are really pleased with how it looks so wanted to share it with you. At the moment it is not sporting PiDrogen team colours but we don’t want to give too much away at this stage!
Things have been quiet on the blog front from team PiDrogen. So, since we are snowed in, what better opportunity for an update?
Well the blog has been quiet because we’ve been coding just as fast as we can. 😀
We’ve developed a peripheral controller based on an Arduino Due. Here’s the breadboard version:
The Due connects to the servos, the SBUS on the radio receiver and to some I2C equipped sensors. It then connects to the Pi via a USB port. The Due updates the Pi several times a second and takes instructions from the Pi to set servo positions etc. Jay, who has recently been press ganged into the team, is busy building a strip board interface to replace the breadboard rat’s nest that you see here.
In addition to this, chief coder Dad has been writing the Pi code. It’s written in Python which he’s new to; so there have been times of slow progress. It basically works (although it’s hard to use unless you attach a keyboard; that’s gonna need work!). Here’s a picture of the app running in calibrate mode:
The image is the view from the Pi camera with the alien outlined in pink by open CV, all at 20 frames per second; pretty impressive for such a small device!
Next for us is to rebuild the chassis. The electronics are turning out to be larger than expected and things are getting a little tight in PiDrogen. Tight fitting electronics and LiPo batteries doesn’t sound like a good combination. 🤔
On Friday, the Team took a trip to the House of Marbles in Bovey Tracey, Devon. The plan? To test the performance of PiDrogen 1’s skid steer, 4 wheel drive chassis against PiDrogen 2’s tracks in a Marble Pit.
Here’s a quick video of how it went….