Thursday, June 21, 2007

Design innovation

On a recent visit to New York, I happened to visit the Marriott Marquis in Time Square and was surprised to see innovation in an area where there has been none for a very long time. I experienced the new technology in elevators.

We have rarely seen any innovation in elevator design, at least from the end user's perspective. There are not too many players in this arena, which in turn is not a recipe for innovation. If any of their clients complained of long wait times, they always came up with other ways of solving the problem. Like, installing flat panel TVs piping News or other programming near elevators to take the mind off the long wait times.

But this new design tries to solve the elevator problem in a unique way. The distinct difference is that there is no Up/Down button to summon an elevator. Also, there is no display atop an elevator signalling where the car is. In place of the Up/Down button is a panel with a number pad and a display. Every elevator has a name (alphabet, in this case). And, the biggest change is that the individual cars have no buttons in them!

Say, you want to go to the 14th floor. You walk into the lobby and key in the number 14 into the pad. The display tells you to go to elevator "E". You go and wait for the car "E". When the car arrives, you just get into it and it delivers you to the 14th floor. The system is trying to schedule the cars and also batch people into cars based on their destination. So, in case of heavy traffic, you don't end up stopping on almost every floor.

For this to work, you need a lot of cars, and a lot of traffic too. The Marriott had about a dozen or more cars. I was not there during the peak hours and hence cannot vouch for the efficiency of the system, but it seems to work.

The one case where it fails is when you enter a floor number (say, 14) and halfway during the journey change your mind (say, 4th floor). There is no way to stop the car as it speeds by the 4th floor. The other case where it fails is when someone rushes into an open car and tries to find keys to punch the floor number. But, both these cases can be considered user errors and dismissed.

The next time you are in Times Square, check it out. To top it all, at the top of the building is the revolving restaurant The View.

Monday, June 11, 2007

Design for usability

Here I go again, bashing another product for poor design. This time it is my digital watch. I have two watches, and I have lost the instruction manual for both. On one of the watches, I managed to set the alarm to 12:00 AM and was unable to cancel it. I tried pressing all combinations of the 4 buttons, but couldn't reset it. I finally gave up and had to remove the battery to reset it. The other watch, I never change the time on it to compensate for daylight savings. I just add an hour for six months of the year. Why? You guessed it. I don't know how to set time on it.

The first watch I got was a mechanical watch with one thumb screw on it. You wound the watch every night by turning the thumb screw clockwise. Whenever the watch ran fast/slow (which it did quite often), you just pull the thumb screw out and rotate it in either direction to set the time. And, once you were done, push it in until it clicks. As simple as that. I did not even get a user manual with it!

And now, we have all these sophisticated watches with chronographs, stopwatches, multi-zone times and all that, and we can't even come up with a simple user interface to it all. Most digital watches have 4 buttons on them. One of them is reserved for light. The other three, in some weird combination, allow you to use/set/reset all the operations. I don't know if any watch maker has figured out a decent and intuitive UI for this.

How would I design a digital watch?

Simple. Create a digital watch with one button, again a rotating thumb wheel. Since this is a digital watch, it comes with a LCD display, and here is how you would use it:
Click the (only) button to summon a UI.
Rotate the thumb wheel to walk through the menu.
Click the button again to select items in the menu.

How much simpler can it be? All the operations can be performed by clicking the button, and rotating the wheel as and when necessary. Borrowing from the old mechanical watch gives us the best UI. Of course, we may end up with some cascading menus, but, hey, you won't need a user manual to use it.

Of course, it would automatically set time based on the WWVB atomic time signal that it catches via radio waves.

Thursday, June 7, 2007

Design for serviceability

Yesterday, the light bulb inside our Kitchenaid wall mounted microwave oven burnt out. Me being a handyman, wanted to replace it myself. I searched for the spare bulb online and found a place selling it. When I tried to replace the bulb is when I realized what I was up against. You need to take apart almost the whole microwave, in order to replace a bulb! First, it needs to be removed out of its enclosure, and then taken apart to get to the bulb. Wow.

I called the service center and they said it could cost anywhere from $75 depending on how long it takes to get to it. Wow. Nearly $100 in labor to replace a $2 bulb (No, actually, the bulb is $20. It is part of a premium product, you see). Talk about serviceability.

I used to replace refills in ball point pens when I was growing up. Later on, these pens became disposable and there was no need to replace the refills. The cost of the pens came down drastically and it made no sense to replace the refills. So, they no longer designed the pens for serviceability, and sealed it shut. Makes sense. It is a disposable product.

When I was working as a design engineer designing material handling equipment (during my Robert Bosch days), my boss used to review my designs. One of the first questions he used to ask me was about serviceability/maintainability. "How would you replace the chain on this conveyor?". If I took more than 30 seconds to explain, or if it involved removing 10 other parts, the design was rejected.

Now, I am looking at an appliance which costs upwards of $3000 (it is an oven combo), and I need to call a service technician to replace a piddly bulb! How wonderful. I thought premium appliances are better designed. Maybe, Viking makes better serviceable appliances...

Good design for serviceability may not necessarily mean a good design for manufacturability. But, in most cases, a good serviceable design is a also a good manufacturable design. The components will be simple and easy to assemble.

On the flip side are the products from Apple. Take iPod for example. For it to be aesthetically and visually pleasing, Apple decided not to put any screws. So, what happens if you want to replace the battery. Tough luck. In these cases, the aesthetic appeal wins over everything else, since that is what this brand represents. This would also mean that the components that go into making an iPod are of high enough quality that they don't warrant regular replacement/repair.

Where does one draw the line? In case of the microwave, nobody cares if there were 4 more screws INSIDE the microwave to hold a receptacle for a light bulb. But, Kitchenaid didn't want that (not to mention the 8- 12 screws right on the front of the enclosure). In most appliances, the design should be such that the fuse and some simple parts (that are failure prone) should be easily accessible. I agree that there is no need to give good access to the magnetron coil in the oven. But, a door hinge, or a door spring, or the fuse, or the bulb. C'mon, they can do better than this.

This shows how much the designers think before they finalize a design. Usability is another big area where Kitchenaid appliances lack big time. Don't even get me started on their cooktop design. That is a topic for another blog.