🛹 Skimmer Boarding: Getting around without the use of your legs and arms

Michael Carney
9 min readJan 6, 2022

Motivation

Ronan is seven. He was born with arthrogryposis, which in his case means he can’t move his arms or legs. He has some neck, hip and torso movement. Ronan got his first electric wheelchair two years ago that he can drive using an adapted joystick and moving his torso to direct it. Electric wheelchairs are great, but they don’t work everywhere. I wrote a little about that here.

What Ronan was missing was something that he could:

  1. Get in and out of himself
  2. Be at the level his brothers are at when playing lego / playmobile etc
  3. Have something small and lightweight enough that we could bring it anywhere with us (without a van).
  4. Have something that was easy and safe for him to drive.

Disclaimer: This is not a design for a medical device.

Note: If you have ideas or thoughts on how to make this better. Or, alternatives you’d like to share, please, please add them in the comments below or mail me at [my name]@gmail.com.

Designing the Skimmer Board

To achieve all the criteria above, and after a lot of research of what’s available in the market and a lot of iterations on possible designs (see appendix), I decided to build something from scratch. Below is the design for the first Skimmer Board. You will see, I have limited talent in this area, just motivation!

Skimmer Board design

Getting the chassis right

The skimmer board is a ~50x90cm plywood platform (1) with two electric motors (3) positioned in the middle of the board. The chassis is angled so that Ronan can shimmy onto the skimmer board himself. When he is in the driving position the platform tilts forward with his weight so that the skimmer drives on the two motorised wheels and two trolley wheels at the front (8). It will run even if it is tilted backwards so that the trolley wheels are not on the floor(e.g. his brother is holding his legs!). The skimmer board is drivable (skimmable?)and doesn’t scratch the floor thanks to the anti-scratch padding (9) and the relatively high torque motors (they are electric window motors).

The board is covered in a soft foam (2) so that it is comfortable for Ronan when he lies on it. This also covers the control panel, making it a lot less painful in the case of a crash, as Ronan drives with his face very close to the control panel and uses his mouth/chin on the joystick (5) to drive.

Controlling the skimmer board

It was tricky to get the motor direction just right from the joystick input and to get the take-off and to tune the acceleration so that the ride wasn’t jerky and Ronan could easily accelerate and decelerate. Initially, I tried to just use the direct control settings on the motor controller (10). Those settings are designed for remote control vehicles and not ones that have a passenger. The sensitivity was way too high even on the lowest setting so I decided that I would instead use the PWM connections and an Arduino (11) to have more control on the motor speed.

This allowed me to set the acceleration, speed and direction and tweak it easily over time. There wasn’t a lot online about how to write speed controls for something like this. I wrote some code that works really well now. Slowly accelerating and moving smoothly between forward, spin in spot, and reverse based on the joystick input. My two main tips are:

  1. Use a function to gradually change the speed on the motors. Joystick input can be a little flaky from time to time. If you don’t do this, you can have a very jerky ride. See the “nextValue” function. I used a 200ms delay on the main loop with 12.5% change in value and this works just right for Ronan.
  2. Break the joystick directions down to the 9 positions in a compass. Again, this leads to a much smoother ride and the gradual changes means you are not just stuck with nine directions. This worked better than mapping functions for me.

Powering it up

I spent a lot of time researching batteries for this. It is really hard to find the right battery and buying a specific battery is a bit of a trapdoor decision that can be an expensive mistake. I was about to take apart an old macbook for the batteries inside when I suddenly realised that I had a big selection of high quality, safe batteries in my shed that were used for powering my drills, multi-tools, saws etc! Playing around with these I discovered that they were perfect. Safe, easy to recharge and come in a lot of different power options. (13) Dewalt were the best brand for my purposes as their batteries can easily be hooked up using standard spade connectors and I have a bunch of them already in my shed.

How did it work out?

It’s ok. I’ll get it myself. — Ronan

This works a dream. Ronan loves it, he’s done miles and miles on it! You can see how manuevourable it is in the video above, and it’s comfy too for him.

I’ve added some additional features over time:

  • Speedometer — this just reads out the -100 to 100 value that is being sent to the motor controller. The kids love this as it lets them know how fast they are going in time trials.
  • Distance sensor — we added this to see if we could prevent collisions, it didn’t work very well and it wasn’t really needed so we removed.
  • Trigger puller — we added a trigger puller so that Ronan could mount a laser gun to the front so he could play laser tag. This was a big hit too.

I’m sharing this as it may be useful for others that might want to build something similar. Or, may have some ideas or suggestions to make this much better.

The skimmer board means Ronan can move around the ground floor of the house super fast and he isn’t left behind when play with friends moves from room to room. It’s light enough to throw in the boot, and low enough that he can get on it himself and he’s at the right level for play.

What’s next

Outdoor version: I built a four wheeler. This didn’t work so well, turning with four independent wheels is way more complicated and/or destroyed the lawn so I need to rethink this. Probably better just to use bigger mid-wheels. The universal dashboard makes prototyping a chassis really easy though, so I can iterate on this over time.

Reliability: Fix the bugs!!!!

This is still very much a prototype made with hobbyist materials. I’d like to make it way more robust so that we don’t have recurring, sporadic issues like loss of control when overheated. I wish I was an electrical engineer and could competently run down wiring design issues, but I’m not, so it’s a lot of intuition and trial and error. If anyone is interested in working with me on a more reliable version of the skimmer board, please let me know! @michaelc on twitter.

Durability: Make it stronger

This is built using wood. Older kids jumping around on it can cause it to, well, snap and break. I would love to move to a lightweight aluminum frame. I’ve started a welding course — watch this space!

Appendix: The process of getting to the skimmer board

I think the key lesson for me with the skimmer board was just to keep trying things and eventually something will work. Below is a quick summary of the process of getting to the skimmer board.

v0 — the green machine

We had played around with some toy ride-ons from go-karts to mini VW Beetles. None really gave him any independence and some downright terrified him. From a practical perspective, there were three major issues.

  1. The driving position wasn’t suitable for Ronan (typically seated) in an unsupported position.
  2. They required steering via a steering wheel which was a non-starter for Ronan.
  3. The turning circle was huge. They worked well outdoors, but not great indoors.

I came across a mid-wheel powered ride-on the Huffy Green Machine at Smyths and this I thought could be a good basis for building something for Ronan. It had a mid-wheel design. It could spin on the spot and could be maneuvered into almost any position. It is lightweight and at a good height making way more enjoyable and engaging for Ronan to play with his brothers and friends.

There were some drawbacks, however.

☹️ The steering didn’t work for Ronan. It was operated by two joysticks that put the right and left wheels in either forward or reverse. He could kinda drive one joystick, but not two independently at the same time.

☹️ The seated driving position wasn’t great for Ronan and still required lifting and strapping him in and out.

Iterations on the green machine

I was moving in the right direction, but it was still not right for any prolonged or regular use. I needed to address the driving position and the steering mechanism. Here is a list of some of the many iterations I tried.

❌ v1.1 — the green machine with connected joysticks. I tried to just join with an elastic band the joystick so that they could kinda be used together. Now, we could go forward, but turning was restricted.

❌ v1.2 — the green machine driven from a laying down position. I stripped out the chair and the control panels and positioned the joystick so that they could be moved by the drivers head. The new driving position was great, but the controls were still not usable.

❌ v1.n — the green machine with assorted switch controls… this was another fail.

A universal control panel

As I went through all these iterations, I had this idea that it would be great if I could build a single universal control panel for Ronan that he could use (it would be adapted for him specifically) and then this could be connected into whatever machine we wanted to drive. Below was the concept.

The hardest part of this design was trying to get a motor controller that would work with the motors that I wanted to drive. After blowing a bunch of L293n motor drivers in spectacular fashion, I opted for this one from Cytron.io.

You can see the latest design in the write-up above.

Green machine v2

The universal dashboard worked really nicely, but the Green Machine motors were pretty low quality and it was really hard to get them to move the machine at a low speed, leading to lots of juddering and not an enjoyable ride. Check out this prototype in action.

Green machine with universal dashboard.

At this stage, I think we can say that we were scoring the following on our criteria:

  • 👎 Get in and out of himself (still too high for him to get on himself)
  • 👍 Small and lightweight enough so that we could bring it anywhere with us (with or without a van).
  • 👍 At the level his brothers were at when playing lego / playmobile etc
  • 👍/👎 Easy for Ronan to maneuver. (He can steer it, but it’s too jerky)

That’s when I decided to build from scratch on a platform specifically designed for purpose.

Other interesting links

I came across some useful sites while researching this:

  • Cytron.io — great for parts and fast shipping from Malaysia
  • Notawheelchair.com — love this project, and gave me some ideas for some future things I could build.
  • GoBabyGo — worth a look at how this charity is helping disabled kids get mobile in the US.

--

--