Servo Magazine (October 2017)
Automate Best Practices
Despite my love of modern technology, I’m old school in many ways. I shave with a straight razor, wear a mechanical watch, use a circa 1940 Singer sewing machine to repair my clothes, and iron my own shirts.
Still, I do my best to keep up with what’s out there, in terms of best practices. You can’t innovate in robotics — or any other field for that matter — if you aren’t familiar with the best methods in current use.
Take ironing. At first glance, it’s a fairly simple, boring task that involves sweeping an iron of the appropriate temperature over a wrinkled garment while avoiding buttons and the front placard. Even so, it’s a task beyond what I’m aware of in current robotics.
For example, the robot iron — designed as part of Columbia University’s impressive Laundroid project (see SERVO, June 2016, p. 17) — can at best iron a plain sheet of material one small area at a time. Identifying and appropriately ironing the collar, sleeves, buttons, and other key structures of a shirt are beyond this or any other ironing robot.
As pointed out in the SERVO article, robotic ironing is a very challenging task. This may not be the last word, however.
My local dry cleaner/tailor uses a small handheld iron for touchups after small mending jobs. For ordinary button-up long sleeve shirts, she uses a hot air ironing system that completely finishes a shirt in less than a minute.
To iron a shirt, she clamps a damp shirt into the device, sealing off the collar, cuffs, front, and bottom. A blast of hot air instantly fills, dries, and smooths the fabric of the shirt, akin to a cloth balloon — all without a hot iron touching and wearing down the fabric.
Now, if I were going to devise an ironing robot for the commercial market, I’d start with one of these hot air ironing systems. Forget the AI and image recognition to identify the wrinkles.
No, I’d develop a system to automate the clamping of the shirt into the ironing machine, as well as a system to remove the shirt and either hang it or fold it appropriately.
It’s the same with any other robotic automation task. Aside from the obvious educational experience, why waste years of effort only to supplant a technology that is no longer current?
If your goal is commercial success, then the bottom line is to know the best practices in the area you’re attempting to automate with robotics. As with my example of ironing, it could be as simple as visiting a local service provider. SV