So it is now just 6 days to go until the Sunday of Pi Wars. This is the last weekend to get things sorted before the big day. So we’ve been frantically 3D Printing and testing.

For the Somewhere Over The Rainbow challenge we were concerned about the way the course was going to be lit. There was no confidence that illumination would be even and that our robot would not cast shadows onto the ball (It’s a pretty tall robot). So we have added some LED lights to the camera side of the robot to hopefully give the camera a better view and to ensure the robot is not thrown off by shadows. We have yet to test this though!

The golf attachment has been through a few revisions as it was found to be very difficult to control in testing. The printer is beavering away right now printing out the new parts.

The Duck Shoot stuff is all working great. Shame it is taking up the majority of the internals of the robot though. Space is tight inside there!

The Minimal Maze challenge is working fine. We are just praying for no bright sun on the day as this will ruin everything as we have not had time to fit any backup sensors.

The straight line speed test works fine. We are not going to be the fastest robot out there (I’ve seen some videos of super fast competitors). However, we are confident we will get from one end to the other without touching the wall so will at least collect some points.

Finally there is the obstacle challenge, which is mainly down to practise as well as Pi Noon.

All in all we have had a bloody good crack at this competition and although we are not fully prepared as we hoped we would be we should be a strong team. Fingers crossed for the day. It will be great fun and we are all looking forward to it. I am personally looking forward to getting my life back past Pi Wars and have the time to get on with some of my own personal projects. The last 8-9 months or so has been dominated by Pi Wars and it will be nice to finally get it done and get things back to normal.

However, I now have some fantastic new skills since starting the Pi Wars competition that will be invaluable for some of my up and coming projects and for that I am very grateful.

Hope to see you all in Cambridge next week. Please come along and support us !

With just 10 days to go till the Sunday that we are competing in Pi Wars, panic is now setting in. We have to admit we are not completely ready and really need another few weeks to get things right. However, we don’t have that luxury. There are 10 days to the competition and considering 6 of those are work days meaning only a few hours in the evening are available during the week, plus once we set off for Cambridge on the Friday any chance to work on the robot after that will be very limited indeed. So this coming weekend is really the last chance to do as much as possible on the robot. After that, time has ran out.

Where did all of that time go? It was way back in 2017 that work started on the robot and thinking we had plenty of time. All of a sudden we are in headless chicken territory and needing more time.

So this is a brief status update on where we are as of today :-

Chassis – Happy with the chassis and we’ve added a little bit of bling to make it look less like an RC car and more like a robot.

Straight Line Speed Test – Not confident about this challenge. Using ToF sensors only so we are weather dependent on this one. No time to add in backup sensors.

Maze – Same as above. Good results in testing but praying for overcast skies on the day.

Somewhere Over The Rainbow. – Robot performs well in testing. Again, lighting conditions on the day that differ drastically from those in testing may be our downfall.

Obstacle Course – This one is entirely down to the skill of the driver. The robot is fast and maneuverable. Hoping to do reasonably well on this one.

Pi Noon – As above. But we are up against formidable opponents so don’t expect to win. Hoping to get at least half way through the rounds though.

Golf – Attachment works reasonably well in testing. Luck and driver skill will need to be on our side.

Duck Shoot – Hoping to do well in this one. We have a nice Nerf Cannon. But then so do other teams so again, speed and luck are required here.

Blogging – We’ve definitely done better than most teams as a lot haven’t even bothered. Hoping for good scores here.

Artistic Merit – Hoping for some points here as we’ve put some effort into making it look good.

Technical Merit – Hundreds of hours have been spent in Fusion 360 and with the 3D printer on the robot. Almost every single plastic part on it was designed from scratch with many iterations in the R&D phase till we had working versions. The Nerf Cannon in particular went through many changes till we got what we wanted. Hopefully high scores here.

So, we are not expecting to win any challenges, but will be trying our damnedest to do so. No matter how we do on the day, I am sure it will still be great fun and I am walking away with some great new skills with Python, OpenCV, Fusion 360 and 3D Printing that will be very useful for up and coming personal projects and those with Medway Makers. It will also be great to put some faces to names of those we have been speaking to on the Pi Wars Discord channel and on Twitter. One thing I’ve loved about the active guys and gals on Discord (not including the ‘lurkers’) is the willingness to share information and help each other, even between competing teams. This sums up the spirit of Pi Wars and the Raspberry Pi community as a whole and I applaud every person involved in this great hobby.

It is now only 35 days to go until the Sunday Pi Wars day and the pressure is mounting. The reality is, we need more time, another 3 months would be great. The reality also is, we don’t have that luxury. Pi Wars is in 35 days whether we are ready or not.

We made some good progress on the Nerf Cannon this weekend and (finally) have a design we are happy with after many iterations. The gun has plenty of power and the addition of the laser sight will help with aiming.  The problem now is finding where to mount it. The chassis is a total mess and needs sorting out. So, I am frantically printing out parts to hang things under and off the sides of the robot to make more space up top to mount the Nerf Cannon and Pi Camera. The attachments for the golf course and the maze sensors will go at either end and around the sides.

We are getting there slowly but things now need to pick up pace. Tom has been busy working away building our test courses whilst Mike cracks on with the 3D printing and coding. Considering it is only a 2 man team, we’ve done really well and have a half decent robot. Looking at what some of the other teams have built it is clear we are up against some very tough competition. As our first entry into Pi Wars as a maker-space we do no expect to  come near the top 10 but will do our best to get as many points as we possibly can. We have certainly learnt a great deal and will be far better equipped to tackle the competition again if we decide to enter another year. Plenty of valuable lessons have been learnt.

However, no matter how we do I am sure it is going to be a great fun day on 22nd April. Even if we don’t win any challenges we’ve learnt loads of new skills and I am now very proficient at building working parts in Fusion 360. The skills I’ve learnt in the Pi Wars build will go a long way towards my own personal ‘Earth Rover’ project which I plan to start once Pi Wars is out of the way later in the year. Tom is picking up tons of Fusion 360 skills which he plans in using to assist with his kids school projects. These are the main rewards of entering a competition like this, the experiences you can take away from it and put to use elsewhere in your life. We are really looking forward to Pi Wars. See you all there!

Over the last few weeks I’ve gone through several design iterations for the Nerf Cannon. This is mainly due to my insistence in trying to use a continuous rotation modded servo to move the darts into position. I have finally given in and used a stepper motor. This is something I should have done right from the start with hindsight.

However, the darts are now lining up perfectly with the gun barrel every time now and the 3D printer is busy beavering away printing out what I hope will be the final version of the gun. Once the gun is finished and fully working I shall post up a video of it in action.

I was saying to Tom yesterday that it is no wonder so many teams drop out of Pi Wars in the first few months. Pi Wars is hard. There is no doubt about it, they are not called “Challenges” lightly. Over the last month or so I have been designing, redesigning, printing and reprinting parts for the Nerf Cannon. We could have gone down the route that lots of other teams have gone down and hacked a Nerf gun, But we chose not to. Those solutions are not elegant, they look ugly as hell, and are not going to win many points for Technical Merit. We chose to design and build a Nerf cannon entirely from scratch.

This was hard for two reasons. One, it meant having to learn a CAD program to enable custom built 3D printed parts to be manufactured. Our chosen software is Fusion 360 as it is extremely powerful and allows entire assemblies of multiple parts to be constructed with relative ease. I have gone from a complete beginner in Fusion 360 to a competent intermediate user in the last few months and I am now producing complicated multi-parts devices.

The recent iteration of the Nerf Cannon for the Duck Shoot challenge is made of 13 separate parts. Secondly, once a design is constructed, there are then rounds of testing, tweaking the design or even complete redesigns, reprinting modified parts and so on. Perhaps if I wasn’t such a perfectionist I would ‘make do’ with an inferior solution, but I am keen to get it right. All the work is made much harder as the bulk of the technical aspects of the robot build fall on my shoulders. Fitting time in to get stuff done is tricky.

For the maze challenge, we have used the usual ToF and ultrasonic sensors and the robot is solving the maze in a very respectable time. This one was a lot easier than expected.

Next, we are going to be working on a solution for the golf challenge. I have a few ideas in my head and parts are on the way. The Obstacle Course and Pi Noon will need minimal work apart from plenty of practice as these are manually controlled challenges. Although we may program in some automation if time allows.

The Straight Line Speed Test will be a lot more challenging with the addition of the narrow section in the centre. This will require either slowing down when it gets there or a very efficient centering algorithm. Somewhere Over The Rainbow may prove challenging, we have yet to see.

Finally, at the moment the robot is ugly. It is a basic chassis with a spaghetti mess of wires on top. I would really like to design a nice looking chassis where all of the electronics are hidden away. However, time is running out and this will have to wait until last. There are only 11 weeks left until the competition. Tick tock tick tock !!!

The 7 challenges for the 2018 Pi Wars challenge differ slightly from before. There are a 3 autonomous challenges and 4 that are remote controlled.

For the remote controlled challenges we have: Slightly deranged golf, Pi noon, The obstacle course and The duck shoot.

Slightly Deranged Golf

The golf course is the same course as last year but with a slightly less steep angle for the initial ramp. This is a tiny golf course but has a ramp to contend with at the start, plus a 180 degree u-bend with 2 hazards, then another ramp with a moving windmill before the hole.

A lot of people had trouble with this challenge last year due to: a) the ramp, balls came loose from holders, etc. Having a way of encapsulating the ball on any angle would be beneficial. b) the hazards, the ball going down the ramp would inevitably go into a hazard unless you were able to aim the ball at a wall and make it bounce off at 90 degrees prior to the u-bend. The ball has to either kept in captivity throughout the course or aimed at speed towards a wall prior to the hazard. c) the artificial grass, this made traction difficult for robots with omni-wheels or small wheels. Go for tracks or larger wheels. d) the narrowness of the course. Several robots were very large and found navigating the bend tricky. Smaller robots fared better here. e) The windmill. The blades of the windmill were turning and so the ball had to be fired at the hole at just the right moment. Some kind of solenoid or other device to propel the ball is advantageous.

Pi Noon

For this challenge you mount a spike on your robot with 3 balloons attached. The aim is to use your spike to pop the opponents balloons. This challenge is won by a combination of luck, driver experience and robot agility. The robot has got to be able to move quickly and do such maneuvers as spinning on the spot and quick reverses. The driver must be well experienced in controlling his robot. Practice, practice, practice. Ideally find someone else with a similar robot and practice against them.

The Obstacle Course

Again, this course is down to driver experience and robot agility. The course is not known until the day of the competition and you get one and one only attempt at the course. The aim is to get through the course as quickly as possible and attempt every obstacle. Points are deducted for rescues or failing an obstacle. Winning robots were not too large, quick and agile with drivers who had obviously practiced well.

The Duck Shoot

The above 3 challenges are the same as last year (apart from the obstacle course being a different course). The Duck Shoot challenge is entirely new. The aim is to fire 5 soft projectiles at 5 duck targets from behind a line. Most people will undoubtedly go for Nerf Darts and either hack an existing Nerf gun or design a firing mechanism from scratch. Again, driver agility and experience is essential here to aim the ‘gun’ at the target. I like this challenge. It is the first time projectiles have been allowed. Coming up with a nice and reliable mechanism that can fire the darts will be an interesting challenge.

Straight Line Speed Test

The straight line speed test is the same course as last year. However, they have added a narrow section halfway down the course. This is to prevent robots simply hugging the left hand wall and to force a more intelligent way of remaining central. There will be an addition of a white line down the centre so either a line sensor or use of a camera and OpenCV will be essential here.

The Minimal Maze

This is the exact same course as last year with no changes. I think only a few robots successfully navigated the course last year so it is a valid challenge. It is essential for the robot to make its way through the maze as quickly as possible without touching the walls as the times for all 3 runs will be added together. Use of infrared and/or ultrasonic sensors will be required here.

Somewhere Over The Rainbow

This is the final challenge and an entirely new one. The robot is placed inside a square arena and has to detect and visit 4 coloured balls in the correct order. There will be now way to do this challenge without a camera and computer vision code (e.g OpenCV). This will be the most challenging of the challenges in my opinion and will require the most advanced code. I think a lot of people will fail this challenge. It will be interesting to see how people fare on the day.

There are an additional 3 challenges where points can be obtained: Blogging, Technical Merit and Artistic Merit. It will be worth devoting some effort in these areas where possible to get a few extra points.

So there you have it. The 7 (10) challenges for Pi Wars 2018. I really like these challenges. They are a combination of difficult and fun. There will need to be some very careful thought and planning on some challenges, especially the line test, rainbow and duck shoot challenges in my opinion. Planning, designing and building for the challenges are already underway. Keep your eye on this blog for details of our progress leading up to April 2018.