With only two weeks to go, final wiring is finish. We have all the sensors we think we need mounted on the robot. It is all coding from here… It is a shame that next post are not going to be judged but I pretty much covered every important part of the building and developing of my very first Pi Wars robot.
We’re going to try to go to a “test” location like Cambridge Makespace to test our robot before competition day to be able to fine tuning the code. No much to do on the hardware other than adding the firing mechanism and maybe another sensor if needed but mostly will be just aesthetics 😀
I was able to build what I think is a reliable bot. With my basic knowledge on coding I was able to use every single sensor mounted on the robot, acquire data and drive it from the joystick.
The As under the sleeve is going to be my amazing partner who will be fine tuning code to be able to cope with all of the obstacles/challenges. I was hoping for him to write a blog on his strategy but work commitments has taken him away from me for the last two weeks and so you’re not going to be able to see his strategy until the D-Day ! Hopefully he will be posting something soon even-though it is not going to be judged. 😦
This blog is about our ideas to solve each of the challenges. We still have a lot to code but we will first concentrate on the autonomous challenges an then the remote controlled ones. Here are the general ideas…
The Hubble telescope nebula challenge: For this challenge we’ll be using the camera to identify the different colours. The general idea is that the robot will make one complete turn around himself to identify where each colour is and then it will be able to follow each one of them in the required order returning each time to its initial location.
The canyons of mars: For this challenge the idea is to try to pre-build the maze so that the robot can localise itself on it. Then, the idea is to use the ToF sensors to map and match the “environment” around the robot while moving. We intend as well to use the cameras and 6DoF sensor to be able to localise the little aliens and make an accurate 90 degrees turn when needed.
Blast Off: the Straight-ish Line Speed Test: For this our first idea was to go to the simplest solution an use infrared sensors to detect the white line and follow it.
I found out a bit later in one of the Robot club sessions at makespace, that there is no obligation to “follow” the line, but to complete the course. There is for sure not much of a challenge using the ultrasonic sensors, but we are aiming to just finishing the course. If we have time, we’ll try to use the camera as well to follow the line and the 6DoF to keep straight-ish and try to increase the speed and finish the course a bit faster.
Pi Noon: This will rely entirety on the skills of the driver (which is not me). The only thing maybe will be to try to add a button on our joystick to reduce torque and speed on the motors to be able to drive it more easily. Otherwise the robot has been equipped successfully with the “famous” connector to mount the balloons.
Same as above for the spirit of curiosity. The Space invaders has been explained in a previous post.
The Apollo obstacle course: Now, I have no idea on how to approach this challenge. Obviously it will rely as well on the skills of the driver… but at least we are sure that the robot will not get stucked as there are nothing underneath the chassis other than the motors. Fingers crossed!
I has taken us a while to think about how to cope with the space invaders challenge. On the Pi Wars website it is recommend the Nerf Darts and firing weapon but we wanted something original and different. My idea at the beginning was to use a device that I have when I was a kid: the Mexican Resortera. It is a really easy to use device as you just only need a Y shaped stick and a rubber band like the ones below. You can find these on any Mexican market or Amazon in the UK.
Unfortunately, this idea would have taken us a considerable amount of time to build with a motorised pulling the rubber band. Kind of the same thing will happen with the Nerf pistol and we are running out of time. My partner is a fan of drones and find out that the parrot Mambo has a mini canyon that look perfect to use on our robot.
The idea now is to do some reverse engineering to be able to mount it on the robot and command it from the Pi Hut game controller. We’ll need to do a couple of modifications on top of the robot but think overall it will be “easy” to use on the space invaders challenge. We have it now and will try to “hack” it, if succeed we’ll be able to participate on that challenge hoping to win some points. Fingers crossed !!
Always be a first rate version of yourself and not a second rate version of someone else.
This is how Nenemeni looks life after (more less) mounting all the elements on it:
I am kind of regretting using “clear” acrylic, as if you “cover” the inside, it has a cleaner look of the bot. (*ordering purple acrylic asap)*. On the above pictures, I didn’t took off the “plastic” on the top layer acrylic and the ToF sensors are mounted on the bottom part of the chassis.
For version 2, sensors will be mounted on the upper part of the chassis, Pi will be rotated and battery and Thunderborg re-arranged. After another night with the laser cutter machine, the results are much better:
The L-shaped front piece to mount IR sensor, connector and ToF was exactly as I expected it. Only drawback: there is no space to mount a third IR sensor if needed. On the picture there is still space for the “hand-made” board for ToF connections on the MUX. Nenemeni V2.0 is looking good.
So Far, there is nothing on the top layer, but hoping to add a HyperPixel to be able to easily launch in between programs for the different challenges. Problem: the HyperPixel uses literally all the GPIO Pins on the Pi, so no possibility to use the IR sensor (which if using the camera, wouldn’t be catastrophic). Luckyly, the Hyperpixel has a breakout I2C connector, so the rest of the sensors are safe… to think about… HyperPixel or Not…
To truly cherish the things that are important to you, you must first discard those that have outlived their purpose.
It is starting to get “messy” on my (3rd) carton chassis, and I don’t need Marie Kondo to tell me it is time for a change. My carton chassis (the third) has for sure lived their purpose, so we need a “serious chassis”. The material: acrylic (if possible purple) 3mm.
One (whole) weekend has been spend on making on CAD (SW) to make sure we’re within the Pi Wars rules. And finally, here it is! Nenemeni is out of the box!!!
This is the V1.0 of Nenemeni from SolidWorks. Next step… the laser cutter and migrate all components to the new chassis. Happily, I have a friend with access to a laser cutter and a bending machine willing to help. I decided to go for the “clear” acrylic (as I already have it).
I wish I have one of those fabulous machines at home (or work). After, the laser cutting day, I could not wait to test if it was as I imagined…
It was almost perfect, honestly… few tweaks and it will be perfect… already fixing some details for Nenemeni 2.0. I do have to say that, the idea from the Tiny 4WD camera holder adapted to Nenemeni’s needs was the right call. I am, in general, really happy with the result… Wait to see V2.0 😀
I Am Thankful For All Of Those Who Said, “No” To Me. It’s Because Of Them I’m Doing It Myself.
That is the question… Of course, the IR sensor do not represent much of a challenge, right?, but it is reliable and at the end, the aim is to go from one end of the course to the other as fast as possible. The question is how fast can it go just with IR sensors? Just in case, I’ll have the possibility to mount a couple of IR sensors on the robot. Maybe, I’ll be able to use both to go “faster”… to think about it…
Once again for IR I took the one from Pimoroni, used on the Bit Buggy. It comes with everything you need to “plug and play”. Just need to mount… having a CAD (carton-aided-design) it is easy to test…
Result of testing, event with the stand-offs that the sensors came with, they are still to high for them to “see” or not the line. For the final design I’ll need to find a way of bring them closer to the ground. When thinking of this… I remember that we need to mount also the “famous” connector for the Pi-Noon and Spirit of Curiosity challenges. I thought that using something like the “camera holder” from the Tiny will easily do the trick. The idea is something like this:
Thoughts, we’ll use the same “front holder” for the ToF, the connector and the IR sensors… *keep thinking….