Has a radical idea ever been the solution to your intractable problem?

12/27/2010 12:33 PM EST

Even at the expense of a project's success, most people would rather mock a wacky idea than use it to spin out a more new concepts on the way to a possible breakthrough.

Way back in the '60s, De Bono said that the important thing about 'left-field' ideas is that they should be expressed - but not criticised immediately. Their value is in stimulating a stream of original thinking, wacky solutions included, during the initial phase of 'brainstorming'. But beware, today most people are still untaught in 'thinking' methodologies, so heaven help the creative thinker in a group discussion.
Write your wacky ideas on notes and pass them to your more creative colleagues, but don't speak them out loud.

12/27/2010 11:09 PM EST

I wonder how many great ideas are not brought to light because of the "that is not how we do it here" attitudes at many companies? The thinking "outside the box" for solutions has been most helpful in my early design career. I was on a small design team with crazy schedule, no resources, timing constraints that could not be met, I think most of you know what I am talking about; anyway, we needed to increase the program memory size for an array processor but could not due to timing. The long and short of it was the available memory devices were not fast enough given our cycle times given the need to have the next program address generated in time for the next cycle. The solution: instead of a program memory address register use a latch, instead of picking the right address generation up-front generate all possible address options (8 different choices) and pick the right one at the end in time to open up the PM address latch. This wrecked havoc on timing analysis but worked wonderfully, enabling us to increase the program memory 4X! It broke a lot of rules but worked.

12/28/2010 12:40 PM EST

@zeeglen, I didn't signpost it but my suggestion about passing suggestions around on notes was realy about presentation skills - if the environment is hostile to innovation, then use tact, get your allies on-board and pick your battles ( actually, covert note-passing is pretty realistic using the i-devices available now).
Of course this nonsense shouldn't be necessary but we don't always work in enlightened companies...

12/30/2010 10:36 PM EST

How many times have departments or individuals pursued a 'new tech' idea to solve a problem and 'almost' had a solution? Keep it real, out of the box ideas can work, but many companies chase them without positive results.

12/31/2010 3:20 PM EST

I work for a company that does engineering for Nuclear power plants. There is not much room for coming up exotic solutions to problems. The template already exists, we just make sure that it keeps working the way it has with the equipment it has since the day the Nuclear Power Plant came on line.


However, I have at times come with design methodology that is very different from how the system was initially designed 30 plus years ago. To my surprise my suggestions didn't go too far. This hasn't stoppped me from coming up with non-traditional ways/techniques of designing the Electrical system for a Nuclear power plant.

1/5/2011 11:36 AM EST

"Thinking outside the box". I've been critical of engineering training at our universities, as much as anyone else has. The old problem of no real world training. However, I've just read about undergrad students in a competition last year that were tasked to come up with an idea that helped the elderly in the kitchen. They actually went and talked to people to find out what their biggest challenges were. It turns out the biggest problem was plugging in and unplugging counter top appliances. The students have developed a two piece magnetic plug. One part stays plugged into the wall socket and the other half attaches to the cord. The plan is for each appliance to have its own "B" half. Very novel and very out of the box.
-BLM

1/5/2011 9:11 PM EST

Not sure if this could be considered 'radical', but at the time was new, different, and solved the problem that had our design team stumped for several weeks. We were trying to phase align a bit clock to the eye center based on resetting a counter at the data transition; this just plain did not work when any transient or white noise was introduced.

The solution was an up/down counter that would only phase shift (advance or retard) the bit clock in a 1/16 bit increment when falling off the upper or lower end of the count. This nicely integrated jittery samples and ignored transients while eventually settling to the bit center. It was a two stage method that used one counter for the back and forth constant hunting, and a second counter to follow only the lowest count (phase) of the hunting counter to eliminate jitter on the final recovered bit clock.

The really crazy thing about this was that the solutions were not generated in meetings at work. Rather, the solutions came out of the blue while I was in the shower one evening after a beer or two. The next day I breadboarded what I had solved in the shower (and hastily written down while still wet before I could forget) and found that this technique worked and solved all out problems.

The moral - when stumped with an intractable problem have a couple beers and take a shower.

1/6/2011 5:20 PM EST

When I was at Neve in Cambridge (audio mixing desks) my senior colleague Steve Barraclough would always make a breakthrough on an intractable problem by picking up the broom and saying 'Time for a clear-up!'
We would take 10-20mins getting the work area spick and span (those desks were 12ft long), maybe fix any worn test leads etc. Suddenly we were fresh and good to go, AND we had a great working environment.
Nowadays the best thing is to get away from the computer screen, I suspect everything you do on a PC goes through the same small set of neurons that handle the interface, no matter what is on the screen, so you get specific fatigue without realising it.
Do you even know where the broom is?

1/12/2011 4:01 PM EST

I have come up with alternative mechanisms to solve production time and safety problems on a few occasions. One time it involved splitting the load/unload process into separate tasks done by separate mechanisms at the same time, so as one part was moved out of the test fixture the next was moved in. That worked very well, the problem had been that the handling mechanism could not move any faster. So I cut the load/unload time in half. Another occasion had a machine to clamp and rotate a brake drum for inspection. The original design had many pinch points, so I came up with a system to lower the drum onto the rotation mechanism with no pinch points. Not only was it safer, but faster and cheaper as well. All this from an EE controls engineer. It was fun.

2/2/2011 3:23 PM EST

One of my first challenges in clinical engineering (the specialty that deals with making everything play well in the hospital) was figuring out what to do about a weird low frequency noise in the Cat scanner instrument room. The room was right next to the patient waiting area, not conducive to soothing nervous patients.

The room was basically a computer room with a raised floor. In the room was a motor-generator unit; it kept power to the CT during a transition from standard to emergency power. This is where my amateur audiophile skills came to work - I figured out that the space between the raised floor and the regular floor were acting as a resonant chamber. The vibration from the M-G unit triggered the resonance, like a finger on a wine glass.

The trick was to fix the problem without shutting down the CT room and redesigning the space. The solution....dampening. I grabbed some 1 inch thick thick foam rubber (industrial floor mats) and we slipped a couple into the open space between the MG unit's legs. The first one reduced the sound immensely, the second eliminated it.

1/5/2012 2:44 PM EST

Don't forget, the 60s was an innovative time at NASA. it was famous for out-of-the-box thinking like that.

I'd like to see similar thinking and culture in today's NASA. Sadly we are using human transport technology that has changed very little since the 60s. Even the private companies that are taking over this business are using same-old same-old techniques.

The greatest achievements in the past 20 years have been observational based. I think it's a truly remarkable feat we have discovered earth-sized planets in habitable zones light years away. It's only a matter of time to discover the extent they hold water like our own. And not a single man was sent in space beyond LEO to accomplish this.

9/18/2012 10:03 AM EDT

One example I developed that comes to mind was a solution to minimize transmitter interference between vehicle mounted radar units. Some scenarios could have multiple vehicles operating in close proximity and the transmitter from one unit could be beaming directly at the receiver in another unit, overwhelming any meager signal received from its own reflected transmitter.

Conventional design called for a tight tolerance timebase to run the microprocessor controlled unit whose transmitter was a harmonic of the timebase and whose synchronous subsampling receiver directly converted the gigahertz received signal to kilohertz audio for further detection processing. The production design used a varactor controlled crystal timbase with a microcontroller driven D/A to continuously and 'semi-randomly' bias the varactor at one of several voltages, causes subtle changes in the timebase, and therefore the transmitter frequency and synchronous receiver clocking.

My 'radical' idea was to use a wide tolerance timebase to drive the timing of the unit, and further, one with horrible temperature sensitivity. A resonator was chosen with specs which statistically lowered the chance of interference between any two radars from something like 1 in 8 to 1 in 5000. Further, the active interference avoidance scheme was changed from one of constant dithering to one of listening during a period where a units transmitter was turned off, and reacting to interference if present. The reaction involved minute heating of the resonator by surface mount resistors located in close proximity to the resonator on the printed circuit board, causing a 'semi-random' slight drift in the operating frequency of the resonator and therefore the radar transmitter and synchronous receiver.

The cost to produce the unit dropped, precious printed circuit board area was gained, and confirmed in testing, the new circuit design outperformed the original circuit design tremendously.