Posts Tagged ‘design’

Where’s my fly-by-wire plumbing?

Sunday, April 13th, 2008

The evolution of user interfaces of “real stuff” is fascinating to me. Some product categories have changed dramatically since I’ve been paying attention. Even common household items that work just fine get updated.

Consider the doorknob. Doorknobs have evolved from a basic utilitarian, mechanical tool to control access to rooms to something much more. Have a look at this example time line (dates are my guesstimate after a bit of googling):

(Sorry, photo lost during upgrade of WordPress.)

(For this post, I’m only interested in the user interface functionality of the knobs — clearly the visual design and mechanical engineering involved evolves.)

In all four examples shown, the basic function is exactly the same: a turnable door knob with a lock. But I’d say that things have gotten more useful — both for people using the knobs and for people installing them. Biometric locks that require no physical key mean less hassle and risk for people, and since knob and lock are now integrated, installation is (I suspect) easier and faster.

We’ve gone from physically turning the lock with a key to a “fly-by-wire” approach, where some electronics are involved. You swipe your key card or fingerprint and an embedded computer tells a motor to do the bulk of the mechanical work. At least in San Francisco, I see plenty of real-world examples of card-entry, and some biometrics, so I know these are commercially available to builders.

So why is there so little fly-by-wire convenience in modern plumbing? There have been some advances — I used a nifty thermostatic shower at a hotel once, which kept the temperature constant automatically to within a few degrees.

Rather than dwell on that question, here is a quick and rough concept I mocked up for a fly-by-wire shower control:

(Sorry, photo lost during upgrade of WordPress.)

Big push-buttons to turn on/off the water, a big analog knob to select a target temperature, a display showing the temperature (so you can know “your temperature” and use anywhere, such as a hotel), a “ready” indicator that lights/beeps when your target is reached, and 3 programmable user presets.

Too complicated? Is this the iDrive of shower controls?

Constraints and the elusive perfect UI

Sunday, April 13th, 2008

User interfaces of all types grab my attention - especially when I feel they could be greatly improved. For example, I snapped a photo of this UI on a machine near my house (Post & Powell in SF Union Square, to be specific):

(Sorry, photo lost during upgrade of WordPress.)

The machine I found this on looks something like this:

(Sorry, photo also lost during upgrade of WordPress.)

My first thought was that only an octopus could use these controls. The system just has no overall metaphor — the closest thing is the vague drawing behind the switches that somewhat correlate each switch to which piece of the machine it controls. It felt not only overly complex, but somewhat dangerous. Actual people would be up in the sky on this thing, and the controls clearly need some training to operate correctly. I didn’t even see an emergency “stop” button or any safety device at all.

Immediately I started mentally designing a simpler, more intuitive, and safer control panel in my head. Fewer controls, each with multi function….joysticks with rotating knobs…safety indicators with audible warnings…etc.

This post was originally going to be about my proposed redesign of this control panel. But I started thinking about constraint. Working with startup companies, I am well aware of how constraints — financial and time constraints, in particular — often impact my own interaction and UI designs. The best designs I envision for my clients have actually never been implemented! Throughout my decade or so of professional interaction and UI design, constraints have always changed the end product.

With this in mind, I reconsidered the cherry picker control panel. I imagine the constraints were typical: budget, timing, people. I guesstimate that there are actually very few off-the-shelf components that would deliver the type of control I wanted to create, so using stock switches is a reasonable compromise. Assembling a panel from stock switches is also much less time intensive than creating custom parts. And I saw people using the machine, and no one fell to their death while using it, so perhaps it’s not as complex as I assumed.

A bit of a ramble, I know…but my take away point is this:

I’m sure that right now, someone out there is looking at UI design work I’ve done, and redesigning it to be better in their head. Whatever design they envision is likely to be cleaner, simpler, and more useful — in theory. But the design I delivered was chosen because I managed project constraints well. Clients keep wanting me to work on their projects, and they are recommending me, so I shouldn’t overlook the importance of understanding how to work within constraints.

Any thoughts on constraints?