Since July of 2001, I have been working on various electrical engineer/design engineer projects, including a heart monitor system for a television game show, a grounding system for a recording studio, a sports/exercise heart monitor, and a cryptographic atomic-decay random number generator. Here are the details for a few of these projects:
On New Year's Eve I got a call from Touchdown Television in New Zealand to work on the ABC Television game show "The Chair", starring John McEnroe.
This game show was based on real-time measurement of the contestant's heartbeat rate; to win you had to answer questions correctly while keeping your heartbeat rate under control.
There was an existing heart monitor system that was designed in New Zealand and had been tested on the New Zealand version of the show. It was really quite good considering how quickly it was produced, but had some serious deficiencies that I had to overcome.
The show was on a very aggressive schedule; on January 1st all we had was a large empty stage, an office suite with everyone sitting on the floor, and a bunch of jet-lagged New Zealanders fresh off the plane. Fifteen days later on January 15th, the first show was taped. We had to work fourteen to sixteen hours per day, but we pulled it off.
In those two weeks, an army of Hollywood technicians built the set, set up cameras, lights, and sound, a computer graphics system to display the heart rate, questions, etc. My part was to set up the heart monitoring electronics.
I started off following the lead of the New Zealand tech crew, but after a month was able to take over the heart rate monitoring department with a newly-hired U.S.crew and thus free up the New Zealand crew to go to the U.K.and set up a show there.
My main accomplishments were running the heart monitoring department so that the New Zealand crew could go to the U.K., getting procedures in place so that we did everything the same way every time, making sure that all the hardware and software was working, and setting up triple-redundant backup systems.
I also identified several new ways to cheat and designed countermeasures, adjusted the procedures to minimize opportunities for human error, worked with the computer graphics crew to resolve any problems translating the heart-rate data to on-screen graphics, and got a good start on designing an improved version of the heart rate monitoring system.
Alas, the improved version was never built. The network moved the show from its original time slot to run against the top-rated "Everyone Loves Raymond" show (with predictable results) then canceled the show.
Another consulting job I have been doing on and off during this time period involves recording studio grounding systems. For many years I have been a consultant to the recording studio industry, removing 60Hz hum and 120Hz buzz from audio systems. My method consists of thoroughly understanding the path that the ground currents take and modifying the grounding systems so as to eliminate noise voltages at sensitive nodes. I also design and install 60/60 volt balanced AC power systems. Balanced AC consists of 60 VAC on "hot" and 60 VAC on "neutral", with the two voltages 180 degrees out of phase. I also balance equipment line filters to minimize induced ground current on the balanced AC system.
Another consulting job was helping an inventor get his idea for a new kind of heart monitoring device patented. I can't reveal too many details because of a confidentiality agreement, but the product hits the market, but it involved packing a lot of microcontroller power into a small package, low power to preserve battery life, and low-cost offshore manufacturing.
Another consulting project was developing the hardware portion of a hardware-based random number generator (HRNG). The customer wanted an "overkill" solution as opposed to the most efficient solution, so their crypto expert decided to use one-time pad cryptography with key distribution via three CD-Rs burned on three systems (Mac/Linux/Windows), delivered by different methods and XOR combined at the destination. For the Random Number generator (RNG) I XOR combined many sources, including multiple HRNGs (low bit-rate devices feeding a cryptographic hash and high bit-rate devices fed in directly), /dev/random, and a bunch of different Pseudo-random number generators (PRNGs), many with random seeds from HRNGs dedicated to that purpose. It is a basic mathematical principle that when you combine multiple data streams with an XOR, if any one of the streams is random, the output will be random, so this method minimizes the chance of obtaining a non-random result. One of the RNGs consisted of a Geiger tube detecting the time delay between radioactive decay events, which is about as random as you can get. I then wrote XOR combining software burned into read-only-memory (ROM) of three dedicated "RS232 combiner" boxes. I had the entire system reviewed by another engineer for possible back doors, and had it all installed the system in a secure facility.