Lessons Learned, MAPLD 2003-7

MAPLD had nothing to do with syrup or trees, despite the usual pronunciation: Maple-dee. It expands to Military-Aerospace Programmable Logic Devices -- but you knew that, right? The PLD part is a generic term for a class of super-chips, of which the most common variety are FPGAs: Field-Programmable Gate Arrays. Say that quickly and it sounds like something to do with doping sports drinks, but no, they're a sort of stem cell among chips, made with kajillions of one-bit storages and simple logic gates, all interconnected in a way that makes them totally useless ... until you program them by impressing electric currents strong enough to break all the connections you don't want, leaving any computer or other logical machine you've designed.

MAPLD itself was an annual conference for engineers working with these devices, sponsored by the Office of Logic Design, a small unit in NASA's Goddard Space Flight Center, because a primary use of these chips is in the unmanned spacecraft Goddard sends to other planets. The leader of OLD is Rich Katz, the hardest-working government employee I've ever seen, and he'd studied enough of the history to realize that many digital logic design questions remain the same decade after decade but each new generation of engineers in the field has to learn them over again. So Rich focused part of the MAPLD conferences on the theme of "lessons learned," and reached out to old rocket scientists (from MIT and elsewhere) to come and share those lessons with the new generation.

I sat with my old Instrumentation Lab division chief Eldon Hall at a presentation in 2003 of the trade-offs involved in applying Triple Modular Redundancy (circuits rigged to do the right thing by majority vote in the presence of a failure). A few minutes into the young engineer's talk, we looked at each other and muttered softly in unison, "My God, nothing has changed."

Eldon and I, and our colleagues Don Eyles and Margaret Hamilton, gave papers at MAPLDs to fulfill Rich's vision; the most spectacular was Eldon disassembling an actual Apollo Guidance Computer on the stage.

We got to know an unexpected shared-interest group who turned up at MAPLDs: hobbyists who build software emulations of the AGC, complete with a visually faithful screen with a working image of a DSKY. Veteran astronaut John Young came to give a talk one year, and Julian Webb of the UK managed to overcome Young's reluctance to autograph anything by showing him an emulation on a laptop. He signed the brushed aluminum case with magic marker, giving Julian the problem of preserving it with varnish. After Young's talk, the MIT Rocket Scientists in attendance were asked to present to him a memento with a genuine Moon rock about the size of a golf ball; as the ham actor of our delegation, I had the speaking part.

Such conferences are largely about networking, and my getting to know Rich led somewhere ...