Are you a tech hero?

Good question Doug!

Back in 2009, I was hired as a firmware engineer at a life science equipment company to help out in a imaging based project.  Being the new guy, I was tasked with a bunch of tasks around the edges of the product, but I noticed a near daily heated conversations about how to improve the image processing algorithms that were developed by a consultant.  I quietly listened, watched and read information on the issues.  Then one weekend, on my own time, I developed a demonstration program, written in Visual Basic, that could process the existing captured image and use simple image processing function to improve the quality of images and also perform proper segmentation of objects in the images.  The following Monday morning, when the daily conversation started, I asked if I could share with them a little program that I developed.  To say the least, they were blown away.  A visual overlay showed that almost every object was extracted from the image.  I was immediately tasked to take over the image processing code development.  It took awhile, but I was able to develop a working algorithm that was able to meet all of the project specifications (accuracy and throughput).  In the process, I developed a patent for utilizing multiple images, at different planes of focus, to improve the accuracy of categorizing object types (which became the basis for additional improvements to the instrument in the areas of improved focusing and improved accuracy of object extraction). 

To make this story even more interesting, to that time, I had never been involved in any form of image processing.  I simply read everything I could find and quickly came up with some simple functions and put them all to together.

Throughout the project, I worked with the junior engineer that had been thrown into the project, and encouraged him to participate in the direction of the development.  We would bounce ideas back and forth, then I would go off and develop code to effect the changes.  We became great friends, as opposed to being rivals.

You are incredibly true. It is what happens very frequently to me. I think that a habit that is unfortunately uncommon in many engineers, is just trying to change the perspective to find the solution. And, eventually (but hopefully), abandon all the academic standardized (by whom?) approach that rarely can be applied to real cases. And I rarely too, I saw working.

Enrico

I would like to call your attention to Dr. Brian Wier a dentist and technology hero who lives in a city near here. At the time of this incident the Dental X-Rays were all film based and each clinic had a processor where an exposed film could be inserted in one end and come out a finished picture on the other end. Most of these machines used a train of gears and rollers to move the films from the developer to the fixer to the wash and finally to the dryer. One day I received a call that Dr. Wier's processor was not working. Unfortunately my schedule did not allow me to get there as quickly as Dr. Weir deemed necessary. When I did make it to the office the machine was functioning so I asked Dr. Wier what he had done to fix it. He told me, "Oh it's not fixed you will have to replace the drive gear on the fixer rack." When I removed the fixer rack I could see that Dr. Wier, using his dental tools and skills had fashioned a functional replacement "TOOTH" on the drive gear. I removed the gear and put on a new one but I kept the gear with the artificial tooth on my desk for many years and Dr. Wier was always one of my inspirations for his ingenuity.

John

In 2008 I acquired a customer in Italy to renew his IT infrastructure including the security layer. In the design of his LAN I included a Cisco Pix firewall, a 15K Euro beast. As the customer is the only one Italian provider of funds data for all the other publishers in the world like Reuters, the system should be up and running 24/7. The Pix firewall was delivered on Thursday with the promise from Cisco that the redundant power supply will be delivered 100% the day after, Friday. Then as always happens, nothing has been delivered until the next Monday.

During the night between Saturday and Sunday, the IT manager called me desperate that the Pix was down, the power supply was crashed. I asked him to open the power supply, check if something was burnt and all seemed ok. The problem was the 2A fuse. Impossible to find at 3 am in the mid of the weekend. The next feed had to start in three hours.

"You are a smoker" I told him.

"Well, open the fuse caps and remove it (burnt) Cut about 2 square cm of your cigarette packet, with the Aluminium foil. Envelop it around the fuse, replace it and restart the firewall"

He, very scared of this operation on the brand new 15K hardware, followed my instructions. The trick is that it works as a fuse and if something occurs, the Aluminium foid heats and the paper burn and cut the foil.

This trick worked. The day after arrived the second power supply, the technician installed it and, three months later, the IT manager told me "Hey, do you remember your cigarette fuse? I remembered to replace it just today!" And int was still in perfect condition.

dougw In the dim past I used to work  (job shop) at Huges Sattelite Systems where it did varied tasks like test equipment repair and then I graduated to sustaining engineering. Most of the time was in support of the high-bay where the birds where.

One day somebody complained that their new wiz-bang spectrum analyzer wound not stay on.  Over the course of the week, we played musical analyzer, from the SATE to my lab, tested it, and back again. So after about 5 days, I went outside to the SATE and tried to figure it out. So armed with my Fluke DVM I checked the power in the cabinets, and they all seemed good. So next I checked the 100amp pull-out on the SATE and that was good. I looked over to where it was plugged in and it seemed ok. But when we power it up again, a few minutes later it crashed again. Fed up I took a stroll following the damn power cable which went behind a movable partition and into the wall. About then I heard a faint click from behind the partition. When I got around the partition I was horrified with what I saw.  My STATE power was plugged into the output of a variac as well as a Weller soldering Iron. This whole mess was then plugged into the wall pull-out. Needless to say, I removed my cable from the mess and plugged it into the wall pull-out directly. It never crashed again! later when I was upstairs in my office I noticed my Boss float by. I relayed what had happed, and suggested that we install a lock on all wall pull-out so this would never happen again.

My billing for that little problem was about 30hrs.

ps that paid for the covers which our machine shop built for us.

A little non-obvious sleuthing and a simple fix saves the day - I hope they showed appreciation.

I am not a fan of AC power problems - I am more into low voltage DC. Lets see …. from the dim past (1986 to be precise) I was installing some big machine exhibits at Expo 86. The power systems on site were a horrible mess - the voltage between the neutral wire and AC ground was 40V and real ground was all over the map. One theatre there blew 40 projectors before they decided to install their own power isolation transformers. My machines were operated by the public, so they had to be protected with GFI breakers - it was a real nightmare trying to figure out how to have them installed so they would perform properly without tripping every few minutes. I didn't get any kudos for that though.

After I had sorted out all my problems I went around looking at what everyone else was doing. One industrial designer was very distraught that her elegant touch screen kiosks were failing every afternoon - they just went totally intermittent working one minute and not the next and back again. Her technical staff could not figure it out, whenever they showed up, the systems worked. I took a close look at the touch screens which turned out to use a fancy optical system for touch sensing (in 1986). I noticed that when a bunch of people were crowding around shading the screen they worked. Anyway I approached her (she didn't know me - I was just some guy wearing construction site attire) and indicated that sunlight was likely saturating the sensors when the sun got to the right angle (in the afternoon) and suggested she install some shade hoods. She was skeptical that it could be that simple but had some nice hoods installed the next day and her kiosks had no further problems. It saved her job, so I let her take all the credit.