So long story short I'm planting a 7.3 IDI diesel behind a 4R100 transmission in a truck and am faced with the prospect of either buying an after-market TCM from Baumann or MSD, or ... ? It occurred to me that with the experience I've accumulated in programming logic it shouldn't be that hard to put together the software side of it, if I have some inputs. Seeing as I have an arduino mega sitting around, I figure I'll try and build it; if it proves too difficult I can buy the after market product. I'm quite motivated to do it as it's a great opportunity to unify a number of skills and solve a very real problem.
Info I can get from the truck with the IDI will be limited to RPM and vehicle speed so I'll be looking to add an improvised throttle position sensor - IDI is dumb as a stump with no computing behind it. I figure between those 3 variables I should be able to get a kind of picture of vehicle load and make a simple logic for shifting decisions, perhaps with an input for weight of load in case of towing to color the logic a bit further.
Am I out to lunch here or is this a viable thing? Anyone have tips or suggestions for this type of project? I'm finding some sample snippets of TCM type code and seeing about what I'd expect in terms of difficulty. If there's any interest, I'll try and document my process for anyone keen.
Any input would be much appreciated! Thanks 
Find another project to sharpen your skills.
Controlling anything that's going on public roads with home built software is asking for trouble, for a variety of reasons.
Why do you need throttle position sensor. You should only need to match engine revs to output shaft revs and have a algorithm that keeps the rev range of the engine in the sweet spot. Something like 1 st gear from 600 to 1500 rpm and tail shaft speed from zero to 100. 2 nd gear from 1000 rpm to 1600 rpm if tail shaft speed between 100 and 250 rpm. Your maths will depend how good the program runs.
wildbill, thanks for your concern - I'll do extensive testing with the wheels in the air and make sure it's good to go. I'm a very cautious driver and have placed safety first as I've been restoring this truck - a complete set of new brake hardware right up to the master cylinder has been one of the first things I've installed. If I were to have F'd up, I'd have a lot more to worry about than the transmission going into limp home. These units are pretty dumb and the logic is not overly complex; no need to overthink it and create disaster scenarios in the mind.
Ripcrow, thanks for your thoughts! My feeling was that without a TPS it'll be difficult to gauge load or downshift on acceleration. How can the transmission know if I'm carrying a ton of gravel or not without the feedback from the TPS? I'm almost certain that's how the stock computers do their math. If I'm revving at 3000 (redline 3500) where it might be okay to shift but the throttle is on the floor and the RPMs aren't changing faster than a certain rate - it should definitely not try to upshift. Just looking at the rate of change in RPMs versus throttle input should do a very good job of indicating load. I'd love to take inputs from the pyrometer and turbo and have a "towing mode" toggle switch to start reading the extra inputs and restricting upshifts when pyro temps are getting high, but that'd be a bit extra ambitious.
mindforge:
wildbill, thanks for your concern - I'll do extensive testing with the wheels in the air
That won't do much good because there will be no load on the system. You need to test it on a rolling-road dynamometer. Or on private ground which the public is not entitled to access.
Have you got written approval for this from your automobile insurer? I doubt they will give it and most places you may not drive on a public road without valid automobile insurance.
...R
Any update on this project?
I had idea to only control the tcc and leave the shifting as it is.
Inputs could be shift solenoids, pwm, coast clutch control