For some reason, Toyota never thought it needed to be available; even the 100 series doesn't have a signal that's exposed (at least I don't think it is). It's only used to trigger a system change; when the ATF temperature gets above the setpoint, the A/C cuts off and an idiot light comes on. Not the most helpful for what I would consider a critical event, but Toyota obviously thought the transmission was robust enough that it didn't merit consideration.
There is no PID, but you can write one. There is a thread in this forum from a year or two back about exactly that; you'll have to search for it. I may have bookmarked it; if I find it, I'll post an update with the link.
EDIT: see next post
The problem is finding the specs for the sensor that's in the cooler circuit. I've been unable to find anything, and I've looked. The service manual has a procedure for checking the sensor, but the data doesn't include the sensor range, that is to say, the output voltage curve compared to the ATF temperature. I think you'd need to know at least two points on that curve to write an accurate PID, and that's is assuming it's linear. If the curve isn't linear, you'd need the equation for the PID. Or you could measure the ATF temperature and output voltage under operation and fit a curve to the data points, and use that to write the PID. That's more work than I have idle time for, at the moment.
My thought was to put another sensor fitting at the point in the line where the OEM sensor is located, either before or after it, the choice is yours. You could use the same fitting that's there now (just couple a second one onto the line next to the existing Y-fitting the sensor is in), and supply your own sensor. That way, you'd have a sensor that you knew something about, and could write a PID for; the problem with this idea is that the output wouldn't be visible to the ECM; you'd have to wire the sensor to an aftermarket display. If you already had a Scangauge, or similar, that wouldn't be a problem. I don't, and I'm not going to have one, so I haven't pursued the idea.
You could also pull the terminals from the existing sensor's connector housing, and replace them with new terminals, after you crimped a second wire onto them. This would be, at most, a 10 minute job. This would provide an alternate output, the same way the previous idea would work, without the mess of having to install a second sensor. The problem is that you'd have the same situation you have now: you wouldn't have the specs for the sensor. But, if your Scangauge, or whatever, was smart enough, it may not be necessary for you to know. I haven't pursued this option for the same reason I didn't pursue the first: I'm not interested in adding a Scangauge.
Both ideas would work for someone who had an aftermarket add-on gauge, I would think.