As you know, Guy, by the 1949-50 and 1950-51 terms, Welchman was at MIT teaching digital computing using Whirlwind as the example computer. A few years ago Forrester said he did not remember clearly why he had Welchman teach 6.602, 6.535, and 6.536, but perhaps it was in order to convince people at MIT that digital computing could replace analog computing (I have the exact quote in an email someplace). The MIT Archive has copies of July 1948 letters relating to Welchman being hired if some opportunity he was seeking in DC didn't work out. I wonder if the Archive has copies of other materials relevant to Welchman coming to MIT (is the Archive still on Covid lockdown?). Greenberg's biography of Welchman may describe what the DC opportunity was; it does discuss his later employment by MITRE. It also has pages about Welchman's MIT team planning algorithms to go on Whirlwind to let it do air traffic control, later expanded to tracking while scanning. On 4/24/2021 2:56 AM, Mark Priestley wrote:
Also possibly of interest is the air traffic control project that they started in 1949 (intriguingly led for a while by Gordon Welchman, of Bletchley Park fame). Perhaps this was in part motivated by the obvious crossover to defence applications. I haven't followed it far enough to see what happened, but they started out investigating how to convert radar data to plane coordinates, and something called "maintaining private line communication" with aircraft. All of which sounds like grist to the interception program's mill.