Project: Wheelchairs
Designers: Pamela Kay Walker - Artist, Activist, Wheelchair User, Disability Specialist, and Author of Moving Over the Edge: Artists with Disabilities Take the Leap
Michael Horton - Software Engineer and general Geek
Ilia Potanin - Engineering Capstone Student, University of California, Davis, 2016
Simon Quan - Engineering Capstone Student, University of California, Davis, 2016
Josh Taggard - Engineering Capstone Student, University of California, Davis, 2016
Matt Seitzler - Mechanical Engineer, Bicycle Geek
Mark Chang - Electrical Engineer, Unicorn Driver
Stuart Robinson - Design Concepts and General Support
Lauren Boswell - Body Worker, Body Mechanics, General Support
Jason Moore, PhD - Mechanical Engineer. Also helped with maintenance on current chair. Lecturer, Mechanical and Aerospace Engineering Department, University of California, Davis
Robbie McMurry - Mechanical Engineer. Also helped with maintenance on current chair
Jacob Taylor - Computer geek, Robotics Aficionado. Helped advise on the initial repairs, and was part of the first brainstorming for WHIM
Dan Fiske - System Administrator and general Geek. Helped with the original design concepts and maintenance on current chair
Dates: 2016
Tools: Wrenches, Soldering irons, Chop saws, Drill presses, Slot cutters, Metal lathes, Keyway broaches, Arbor presses, Circular saws, Squares
Parts: Frames, Nuts, Bolts, Motors, Batteries, Wires, Electrical connectors, Wheels, Axial bearings, Shaft collars, Wheel hubs, Casters
Techniques: Wheels and axles, Tri joints, Bolting, Caster joints, Dependent suspensions
Licenses: CC BY-NC-SA


A wheelchair is a chair with wheels, used when walking is difficult or impossible due to illness, injury, problems related to old age, or disability.


Current electric wheelchair users constantly have to deal with high insurance prices and long wait times when it comes to purchasing and repairing their electric wheelchair. Specifically, our sponsor Pamela had to wait several months just to have her caster wheel assembly replaced. For someone who requires their chair for a main mode of transport, this is unacceptable. Additionally, the wheelchair manufacturers have been creating chairs with lower specs (torque, average speed, etc.) while insurance companies are raising the costs to own such a chair.

The goal of the project was to avoid these negative setbacks and also prove someone could build their own electric powered wheelchair in their garage using basic tools such as a chop saw and drill press. The scope of our project included producing a working drive-train and frame that would provide a solid foundation for our sponsors, WHIM Unlimited, to develop a fully working and refined electric powered wheelchair. We hope this assembly guide will provide the necessary framework for others to build their own chair, ending the longtime dependence on inconvenient health and insurance companies. This guide is by no means includes a final product, but rather a proof of concept and a framework for others to build off of to fit their own personal mobility needs.

Based on our meetings with our sponsors, we determined our chair would need to exhibit the following:

  • Powertrain
    • Powerful motor
    • Ability to climb road curbs
    • Good acceleration
  • Body
    • Stable and sturdy
    • Maneuverable
    • Meets ADA specifications
  • Overall
    • Safe
    • Modular
    • Easy to build (no complex geometries)
    • Easily obtainable standard parts (minimize custom parts)
    • Long lasting
    • Adaptable to the user

Wheelchair maintenance is problematic, at best. Parts are available from limited sources, usually medical supply specialists. If you're using insurance to get the chair fixed, then often a prescription is required even for simple fixes (including flats!?).

That means, if something breaks, you need to

  • Schedule an appointment with your doctor to get a prescription (1-3 weeks)
  • Wait for the prescription to be approved (1-8 weeks, depending on type of repair)
  • Schedule an appointment with the repair shop (1-3 weeks)
  • Schedule a van ride to the shop
  • Wait in the shop while they analyze the problem (several hours, with no access to your chair)
  • If they don't have the part,
    • Go back home while they order the part (days, weeks, even months)
    • Schedule another appointment with the repair shop and the van once the part comes in (another 1-3 weeks)
    • Wait in the shop again while they fix the problem (several more hours sitting in their waiting room, still without access to a functioning chair)
  • And finally head back home with a working chair. Maybe.

In our experience, this can take up to 2 months or more, depending on the issue. It can be worse if there are no local wheelchair repair shops.

If you're not using insurance, it might be quicker, however the costs associated can be quite high. US$10,000+ for the initial chair, $200+ for a single battery (you generally need 2), $400 for an seat cushion, and so on.


Based on our project needs, we created the following project specifications:

  • Powertrain Specifications
    • 3 MPH up a 4 degree incline
    • Can go up a 10 degree curb (bump)
    • Accelerate from 0 to 5 MPH in under 2 seconds
    • Max speed of 8 MPH (wheelchairs on the market average about 4 MPH)
  • Frame Specifications
    • Max size of 30"x42" (ADA)
    • Chair height of 15" to 25" off the ground
    • Can handle a 500 lb load
    • Turn radius < 44" (taken from ADA chair dimensions)

Bill of Materials

As stated previously, we wanted our materials to be readily available and easily accessible for anyone to find. The reasoning behind this was to avoid long downtimes or having to go through a supplier that has a monopoly on selling specific parts (i.e. only one set of motors that works and only one supplier who offers them). This would also allow someone with a chair in need of repair to go to their local hardware store or order a part online and replace the faulty one quickly on the chair themselves or with a little help. Thus long down times would ideally be eliminated.

As part of our project, we researched different suppliers to find our initial sources for the parts, but we also found alternate suppliers and parts in case the primary source disappeared. The overall cost of the chair was roughly $2000 as compared to chairs on the market that range from $1000 to $6000+. Our chair is on the lower end of this price range, but it offers the specifications one would find on a more expensive model. Overall, our cost could be reduced with various design improvements and material changes, but again this was meant to serve as an initial prototype and proof of concept.

Tools Needed, Custom Parts, and Gridbeam Cut List

As Mechanically Engineering students, we had access to a full machine shop capable of creating all the custom parts to our hearts content. Unfortunately, the average person doesn't have this luxury, thus if we took advantage of these tools then it could get quite difficult for someone to follow our instructions. For this reason, we chose to work with tools that the average person may have laying around in their garage.

A majority of our build was made using only a chop saw and drill press, however we did have to make a few custom parts. These included the driveshaft, motor mounts, and driveshaft coupler parts. We felt these parts required the most stability and reliability possible, so we chose to make simplified custom parts. We then created drawings for these parts so someone could go to a local machine shop or makerspace or even contact an online service to have them made.

We were able to use only a few custom parts thanks to our choice of framing material: Gridbeam. This material is best described as 2" square tubing with 7/16" mounting holes every 1". Unfortunately, the material is fairly heavy (~5 lbs/foot) but it does make assembly fairly easy once the cuts have been made.

Caster Assembly

During the summer before my Senior year, I had the privilege to work at Aerojet Rocketdyne as Visual Planning Intern. Long story short, we would take advanced rocket assemblies and create a virtual build of the model so the technicians on the floors had something easy to follow when assembling the parts. Basically, I created Ikea/Lego-like instructions for rockets. Seeing as the wheelchair needed a set of easy to follow plans, I created a guide by applying the same skills I used during my internship. The finished result is a clean, easy to follow list of steps that doesn't require a B.S. in Mechanical Engineering to follow.

We created our entire assembly using Solidworks. I then planned out the best way to assemble the parts, and create the virtual build by putting together vector drawings using Adobe Illustrator and InDesign. The first assembly we decided to build is the caster assembly. The caster wheels make it a little awkward since they rolled around fairly easily, so it's best to have some help or somewhere to put the assembly as it comes together. Make sure all the Gridbeam pieces are square to each other before tightening the bolts.

Motor Assembly

The next assembly to put together is the motor assembly. This one is a little trickier to put together and required some finessing in order to get the motors, driveshaft and motor mounts to all line up. Luckily none of the fasteners are permanent so if you mess up there's nothing to be worried about.

Final Assembly

Once the caster and motor assemblies are together it's time to put it all together. Both assemblies are fairly heavy and awkward in size so it helps to have at least two people to put them together.

Controlling the Chair

Our chair utilizes the open sourced Arduino and two H-Bridge shields to control the electricity to each motor. We also use a USB host shield and a cheap USB joystick to provide an intuitive control scheme. For batteries, we used two 12V lead acid batteries hooked up in series to provide a total voltage of 24V. We found this to be more than enough power and control for the prototype, however the Aruidno code needed to be refined further before the chair could be used for everyday use.

Note: I mainly worked on the overall chair design and documenting how we assembled the build. One of my other teammates worked on the actual Arduino code, but I believe he just researched code that other people had used with the H-Bridge shields. Again, we were Mechanical Engineers so learning/writing code from the ground up wasn't really really in our scope. Apologies if the lack of refined controller code is what prevents you from making your own chair.

Testing the Chair

As part of our project, we needed to prove our chair met the specifications we outlined at the beginning of our project. We also participated in a Senior Design Showcase put on by our school where all our classmates also showed off what they had worked on for two quarters. Everyone at the showcase was impressed with the maneuverability and ruggedness of our chair and we didn't shy away from showing it off. Attached are some videos of our initial benchmark tests along with footage from our showcase. By the end of our project, we were happy to have something that actually worked and had a fun time showing it off and playing with it.

Design Recommendations, Add-ons and Accessories

Seeing as our course was only two quarters and we had a finite budget, we didn't get to fully develop the chair like we would have. It's obviously missing a chair and other elements to provide a smooth and comfortable ride. We iterated through our design countless times but there's always room for improvement. Again, this is mainly a proof of concept and initial crack at a problem, so there are many things that could improve the chair.


  • Headlights, tail lights and running lights should be standard!
    • LED or better technology? (Low-power, long-life)
    • Lights should be set up such that they will NOT run if the battery is below a certain percentage
      • Don't let the chair become useless because the lights were left on!

Signals and other stuff

  • Turn indicator, and/or brake lights?
  • Reverse lights or 'beeper'?
  • Rear-view camera?

The Seat

Possible Sources

  • A basic plywood seat, made from parts found at your local hardware and fabric/craft stores
    • Optional: Padded at your local upholsterers shop
  • A chair from your local furniture store or IKEA, modified to fit the frame, padded as necessary
  • A seat from a car or golf cart
  • A modified office chair, such as an Aeron (http://hermanmiller.com/Products/Aeron-Chairs)

Certainly the seat cushions can come from your local upholsterer, if you can't find a source for pre-made cushions.

Possible Features

It would be nice if the seat were

  • Adjustable
    • Reclining (powered or manual)
    • Sliding (forward and back)
      • Move forward for easier transfer
      • More stability on hills (able to shift center of balance as needed)
      • Possibly done with heavy-duty drawer glides, mounted to the frame
    • Folds forward
      • Fit in most SUVs and some trunks
      • Similar to front seat of 2-door cars
    • Pivoting?
      • Allow access in conditions where maneuvering is tight, such as sinks in small bathrooms
    • LOCKING for all of the above
      • Must be safe and stable while traveling!
  • Repairable
    • Easily repaired individual components (seat, back, arms)
  • Replaceable

Dream features

  • How about a drawer underneath the seat, for documents, tools, etc.
  • And another storage area behind the seat-back
  • A lift would be nice, to raise and lower just a couple inches
    • Possibly using compressed air?
  • Self-leveling, for when you're parked on a slope (such as a movie theater)
    • Forward-backward at a minimum, but left-right would be nice, too.


Possibly the easiest component - a strip of plywood cut to the individual's specifications, padded and covered. Use wood screws or T-nuts to connect them to the frame.

Possible Features

  • Perhaps hinged, with storage space underneath for wallet, pens, etc. (car center console)
  • In-arm tray (airplane dining tray, or classroom desk)
  • Built-in coin sorter (okay, a bit weird, I know)


Recognizing that different users may have unique requirements, this will probably be individualized for each rider. Some possible configurations are:

  • No foot rest
  • One, center-mounted, swinging up when not in use (could be used to carry groceries!)
  • One or Two, side-mounted, swinging out
  • Flip-up, Barcalounger (recliner) style

I'm still trying to figure off-the-shelf sources for some of these, and would love some feedback.

Main Brakes

This is another area that is giving me problems. What off-the-shelf solutions are there for normally-closed, electrically-opened, low-power, manually-clutched braking systems?


When there is no power being applied, the main brakes should be engaged
The brakes need to be operated electrically
The power required to open the brakes, and to keep them open, should be minimal
There must be a way of releasing the brakes manually, so the chair can be pushed

Auxiliary Brakes

And even here, I have yet to find a simple off-the-shelf solution for some backup brakes which can be triggered manually.

Current Motor Specifications

I've pulled the motors from Pamela's old chair. According to the plaque, the specs are:

  • 24V
  • 6A
  • 2400rpm
  • needs to be capable of at least 7mph with a 10 mi mininmum range
  • Torque: Unknown
  • Gear Ratio: Unknown, but from some numbers we're run, I'm guessing 20:1 or 24:1
  • Braking: Magnetic brakes, normally closed
    • When power is applied, an electromagnet opens the brakes
    • When no power is applied, a spring closes the brakes
      • A manual clutch can disengage the brakes entirely
  • Doing the math, each motor is approximately 1/5 HP (Thank you Mark!)

Motor Requirements

  • Bi-directional
  • Motor Braking
    • Can short the motor to itself to limit the speed, but this will not stop movement entirely
    • Could use active braking (controller senses movement, and actively moves in the opposite direction)
      • Wouldn't work if the battery were drained
      • Would require more power (controller would have to remain powered, motor might remain engaged at low power for long periods)
    • Could try to replicate the current electro-magnet option, but from what standard parts?
  • Clutch to allow for manual operation of the chair

Motor Questions

  • Which type of motor
    • "Standard" 12 or 24v motor
      • +Lower RPM, so not as much gearing
      • -More expensive ($500+ from most sources)
      • +Power closer to what we're used to supplying (24VDC, 6A)
      • -Low horsepower (0.2HP)
    • Treadmill motor
      • +Cheap (less than $50 on EBay)
      • ?High RPMs
        • Possibly lowered enough via power management
      • -High voltage (~120VDC) and Amperage (20A)
        • Need to play with this using smaller values, to see how it performs, and what the current draw is like
      • +Much higher horsepower (~2.5HP)
    • For either motor type, can the output shaft support the weight of the chair
      • Or do we need to use a pulley (or chair, or direct gear, or...) to drive the wheels
  • How do we stop this thing without requiring a constant power drain?
    • Imagine, looking down a hill in San Francisco, with a low battery, needing to get home...
  • Then, how do we release that brake manually so the chair can be moved without battery power?
    • Now, can we make it so that the driver can control the clutch while seated in the chair?

Motor Assumptions

Desired Speed -> RPMs
mph in/min RPMs (26”) RPMs (24”) RPMs (20”)
10 10560 129.28 140.06 168.07
9 9504 116.35 126.05 151.26
8 8448 103.43 112.05 134.45
7 7392 90.50 98.04 117.65
6 6336 77.57 84.03 100.84
5 5280 64.64 70.03 84.03

Chair Requirements

  • Drive wheels using standard bicycle sizes
    • 26" = 81.68" circumference (depending on tire?)
      • Very common
      • Probably too big
    • 24" = 75.40" circumference ...
    • 20" = 62.83" c... (BMX bike wheel)
      • Very common
      • Often used for heavy-duty riding (tricks/stunts)
        • Are they more durable?
  • Two motors (1 per side)
    • An alternative would be four-wheel drive, using Mecanum wheels
      • Which would be great if you only drove on nice smooth surfaces
  • Target top speed of 10 mph (16 kph)
    • Ability to limit this, based on user requirements, normally limited to 7 mph
      • Allow for "Turbo", for those long boring straight stretches?

Weight Requirements

Completely arbitrary for now...

  • 200 lbs Rider
  • 250 Chair (Frame, batteries, motors)
  • 50 Accessories (Customization, lights, computer, ???)
  • 200 Cargo (Groceries, towing a manual chair or trailer)
  • 200 Passengers (Grandchild on the lap, nephew on the back...)
  • 100 Just in case

Total planned weight capacity: 1000 lbs (450 kg)

Maintenance Guide

No machine works well just by itself, but instead requires maintenance to prolong it's life. Our wheelchair is no exception. Although we were done with the project a few weeks after it was built, we still were able to forecast which components would require upkeep in order to have the best performance possible. These suggestions are mainly based on the manufacturer's suggestions for the parts, and we don't know for sure if problems would arise with other parts.

Assembly Document

Overall, this project was my first time applying my coursework to an actual, real world problem. I hope if you choose to follow this Instructable that you share with me your hardships or any suggestions for improvements you may have. At the start of this project, we wanted to make the chair easy for anyone to follow, but at the same time allow others to adapt it to their own needs. Thank you for taking the time to read through my Instructable, and I hope you enjoyed it as much as I did making it and the DIY Electric Wheelchair contained herein.



There are two types of wheels we need to address: Drive wheels (connected to the motor) and Caster wheels (swiveling wheels used to maintain stability).

Using what I vaguely remember from railroad wheel configuration (Whyte) notation (front casters - drive wheels - rear casters), the layouts I can think of are:

  • 2-2-0 (Old traditional, rear wheel drive)
    • Requires tip-bars to keep from popping wheelies
      • And this is a negative?
  • 0-2-2 (New traditional, front wheel drive)
    • Said to be less stable at higher speeds
  • 2-2-2 (Six wheels, drive wheels in the center)
  • 1-2-0 (Triangular configuration, often used in scooters)
    • More prone to tipping?
  • 0-2-1 (Reverse triangle, single caster in the rear)
  • 1-2-1 (Diamond configuration)
    • I've never seen one, but it's a possibility
  • 0-2-0 (Only two wheels - self balancing)
    • Segway PT
    • iBot
      • Gotta ask, though - what happens when you run out of power, and how do you transfer?
  • 0-2-2-0 (Four-wheel drive, utilizing 4 motors)
  • ?-2-? (Not sure how to classify this one)
    • Tank Chair
    • Rip Chair (As designed, a manual chair carrier, not a power chair)
    • Very cool, but not sure I'd trust them around the house, or even in my own back yard!

Wheel Types

Drive Wheels
Large vs. Small

...Coming soon...

Weight Limits

How much can the mountain bike wheel support, for constant use? Same for any of the other types. Our current estimates for wheelchair minimum weights are:

  • 100 lbs 2x 12V ~75AH Gel Cell Batteries)
  • 50 2x motors (guesstimate)
  • 50 Frame
  • 100 Rider (I did say minimum)

We're already at 300 lbs (135 kg), and we're not even considering cargo, heavier drivers, the extra wear caused by rough terrain, or anything else.

Connecting to the Motor

The obvious possibilities here are:

  • Chain drive
    • Standard technology, used in bicycles
    • Dangerous, especially in motor-driven applications (easy to get clothing, sticks, fingers, etc. caught in the gears)
    • Noisy? (Not as certain that this is still true, but that's Pamela's recollection of past chain-driven chairs)
    • Less responsive? (Slack in the chain could lead to momentary loss of control when power is applied/removed/reversed)
  • Direct-connect
  • Hub-motor
    • The motor is contained within / part of the wheel
    • Uncertain if bi-directional (requirement for a wheelchair to be able to turn and reverse)
    • Uncertain on torque and speed
      • Smaller motor usually equates to weaker - can a pair of hub motors get a heavy wheelchair frame, batteries, driver and cargo up a steep hill safely?

Connecting to the Frame

IF we can use the motor as the connection point, we need to worry about whether the wheel can withstand the stress of being supported on a single side (cantilevering). Traditional bicycle forks support the wheel on both sides equally. If we use a traditional fork configuration, though, it potentially complicates the connection to the motor. There are a few single-blade fork designs.

Another alternative would be to build a removable cage around the wheel - providing support and protection, but also adding weight, complexity, and possibly width, while removing easy access for maintenance.


  • USB charger for your phone or MP3 player
  • Reading light, possibly USB powered?
  • Fan USB again?
  • Drop-down deck on the back for passengers
  • Cup/mug holder
  • Dinner / work tray
    • Re-purposed from an airline in-arm dinner tray, or a school desk with flip-up arm
  • A drawer under the seat for papers, books, tools, etc.
  • Side compartments for purse, wallet, keys, coins, phone, pens, etc.
    • Mounted under the armrests, like a car's center console
    • Or, opening in to keep unwanted hands out
  • Trailer hitch, possibly for a bicycle trailer
    • Groceries
    • Extra batteries or other power sources (solar, fuel cell, etc.)
    • Preferably attachable/detachable while seated
  • Clip for umbrellas, walking-sticks, canes, reachers and walkers
  • A hydration system, such as a Camelbak Pack