It fits like a glove: The UX of hardware prototyping
Want to skip to the video? It’s here.
It seemed simple at first. The idea was to make an Electroluminescent Light-based signaling system. Some switches, some lights, some wire and we’d be done, right?
As usual, we challenged ourselves to produce a practical wearable product. What can we produce that’s new, interesting, and useful? A common scenario: Short time spans and high demands.
Even for internal exercises we need to flesh out the concept. We went through the process that we go through whenever we’re presented with a new challenge: Who is it for? How will it be used? What is it supposed to do? What problem will it solve?
For software, there are many tools and techniques for user testing and experience validation. It’s trivial to collect data points on a user’s behavior, correlate that with other sources, and build a pretty report on the recommended course of action.
It’s not as simple for hardware. Each interaction from every user will be different than another to some degree. It’s the nature of the physical world. Additionally, in the prototyping stage, it’s difficult to iterate as quickly as you can with software. It’s much easier to swap out some lines of code than a knob or switch. The decisions have to be made, primarily, off of qualitative data.
Let’s walk through our process, from concept to prototype for this product…
🧠 Round One: Concepts
Our original idea was to make a signaling system for bicyclists. That meant the controller needed to be lightweight, easy to operate and unobtrusive. The indicators needed to be simple, clear, and useful to both the rider and the people they share the road with. The whole system needs to be portable and adaptable to any bike of any size.
An insight: As silly as it might look or feel, “Bodystorming” is a great way to prototype physical interactions. Microsoft has some great insight on this process. Obviously, it’s imperative that you keep the user top-of-mind throughout the entire design process.
Our initial thoughts for the controller revolved around attaching something to the brake handles. But we had concerns about the portability and obtrusiveness of that installation. What about a hand mounted system? What if the controls were your fingers?
Always a fan of the Nintendo Power Glove, it served as our original inspiration. As something more practical for a bicyclist, we chose a generic ‘cycling glove’ off of Amazon and went with it.
For the indicators, we wanted to use Electroluminescent Paint to create something like what you see in ‘festival masks’. However, sourcing the ink in time and at a reasonable cost was difficult, so we went with the old standby EL tape.
We settled on a backpack as a mechanism to mount the tape to, and as a container for the various electronics needed to run the system. It seemed feasible to us, that were this productionized, it could be reduced to a size that could be placed in/on a real backpack.
To control the EL elements, we needed to use a relay board. This allows us to isolate the low-voltage DC that controlled the arduino from the High Voltage AC that powered the lights.
⚙️ Round Two: Prototyping the prototype
The glove fit well and offered many different methods to attach complementary components to control the lights. We started with magnetic switches, but they were too bulky. Alternative ideas revolved around ditching the glove and switching to some type of inertia tracking device or vibration sensors connected to various places on the bicycle. However, nothing was as simple and direct as manual manipulation.
We settled on buttons under the fingers. There’s several different types of switches that we could have used, and even more models of each of the different types. But through testing we found small, simple momentary push-button switches to be the easiest and most reliable option.
A discovery: Grab a pen in the same way you would grab a bike’s handlebar. You can feel that the majority of the pressure isn’t under your fingertips as you might assume at first. It’s actually under your finger's first and second set of joints. We needed to mount the buttons up towards the middle of the glove’s fingers to ensure maximum pressure and control on those buttons. Additionally, we mounted the buttons on the inside of the gloves to allow for a better feel of when the buttons were being pressed.
Well, that’s it. Press some buttons, turn on some lights… cool, I guess? Not yet, we needed more…
🔥 The Pivot: What’s hot in the streets?
It was around this time that Chicago launched their e-scooter pilot. And with that came all the controversy and intrigue about the new two-wheeled menaces cluttering our sidewalks and bike paths. The scooters themselves are IoT devices. Tracked, controlled, and activated through various combinations of bluetooth and cellular communication.
A discovery: Bicyclists already have an established hand signal system to indicate turning, stopping, etc… Scooters, on the other hand, must be controlled by two hands. The slightest accidental movement at speed will send you tumbling down into the asphalt. Our single-handed indicator system could solve two problems at once.
For this to be timely, we pivoted to envisioning this as a scooter indicator system. Additionally, we needed to spice it up with some Bluetooth Low Energy communication.
🥶 Feeling a little blue
Wiring directly from the glove to the backpack worked, but it wasn’t very cool. Luckily, we had some spare BLE-enabled devices laying around and decided to put them to use…
After laying out the prototype for the BLE interface, we quickly realized a key opportunity to enhance the physical UX for the rider. Instead of pressing the buttons to turn on the signals, what if the rider triggered the signals by lifting a finger? This would automatically turn on the brake light whenever pressure was released from the throttle.
To test this hypothesis, we included a switch on the protoboard which changed the switch values from NO (normally open) to NC (normally closed) and vice-versa.
A discovery: In code, you can implement an A/B test with a simple if statement. However, in hardware it’s a little more difficult. Consider ways to test implementations with spare input or output mechanisms that affect the behavior of the device. Obviously, you don’t want to introduce any variables that could affect the test.
A second discovery was that if the rider is riding with the indicators on their back, it’s hard to validate if they’re working correctly. A new technique was introduced to our repertoire: Audio debugging. We placed a piezo buzzer on the protoboard which would change it’s sound depending on which buttons were being pressed, if any.
🛴💨 Let’s Roll!
Dan Thompson is Director of Emerging Experiences at Perficient Digital Labs. Read more about how Perficient Digital Labs uses Emerging Technology to create innovative experiences that transform the way our clients do business here.
In an experiment, most links on this post are Amazon Associate links. As an Amazon Associate I earn from qualifying purchases from those paid links.
