Just thought I would post an update on my son's Panda 1.2 Dynamic dualogic which had all these faults when we bought it - post
#562 in this thread. It may be tempting fate to say so but it seems that we may have finally managed to resolve the issues as it has now been driving without problems for the last 3 months - even if somewhat less than usual due to the covid situation

Previously we were getting frequent failures and at least one on any long trip.
To recap the Panda was; dropping into neutral at lights, stalling frequently, sticking in gear on the motorway and then taking 20 minutes before resetting iteslf. Here's a short resume of what we did;
Firstly we spent ages going over the gearbox readouts from the AlfaOBD and
multiecuscan applications trying to figure out what was going wrong. Although we got lots of seemingly useful information there was nothing there which pointed to any specific solution. Cost to this point was 50 euros for a
multiecuscan update to run on an ipad (it has worked fine on a laptop previously, didn't work properly on an iphone but seemed to do everything ok running on an ipad apart from resettiing the clutch point but in the end that didn't matter).
The stalling was probably the most annoying fault due to the frequency of it happening so we went for that to start with. There was some suspiscion that the clutch may have been out of adjustment, especially since multiecucan adjustment routine wasn't working and I had also noticed a slight leakage of fluid through the clutch position sensor. Given that I have a second robot unit under the bench in the garage I swapped over the clutch position sensors and did the end of line calibration of the gearbox using the software. There was no improvment but thankfully no more expense either.
Next we considered it could be something with the engine iteself, first cleaning the throttle body (no change), changing the MAP sensor (no improvment) and finally the crank sensor (result!). No more stalling and cost £25 for the sensors from ebay.
Now we came to fixing the real gearbox issues it was hard to know where to start. The softwares both seemed to be telling us that all the sensors were working correctly and that for most of the time the pump was running and pressurising the system as it should. On the occasions we were both in the car and could monitor things as we were driving the only noticeable thing was that if the hydraulic pressure dropped below 38 bar the gears would not change again until the pressure had very slowly bled away down to zero. It was clear that the pump was not restarting but given we knew the pump was otherwise working it seemed an unlikely candidate and far too expensive to change on speculation without more evidence.
On one occasion, giving the robot a sharp tap on top with a spanner (17mm ring I think it was) made the pump restart and got us moving again. but other times this tactic did not work so I'm not recommending it to anyone here. As a result of this I also tried swapping out the pressure sensor on the top of the robot but with no effect.
Renewing the hydraulic fluid (Tutela from ebay £20) seemed to improve things for a while but not completely.
Warning THIS NEXT BIT COULD BE DANGEROUS AND CAUSE YOUR CAR TO CATCH FIRE IF YOU DON'T ABSOLUTELY KNOW WHAT YOU ARE DOING! To get a better handle on what was going wrong with the pump I fitted a long trailling lead to the pump relay in the engine fuse box. If you don't know which relay it is you can listen for it clicking as someone opens the driver's door. The trailing lead was pushed into one of the coil (small) terminals on the relay and the was long enough to touch on to the positive battery terminal thereby operating the relay and starting the pump thus confirming that both the relay and the pump were working correctly.
Taking extreme care that the end of the lead could not short out to earth, I coiled it up and left it in situ with instructions to my son to use it next time the gearbox locked up.
Sure enough it was not long before I got an elated call saying that the box had failed and that the trailing lead thing had restarted it. The question now was why??? It took us a while to ponder this, all the while using the lead to unstick the gearbox until one day I got the not so elated call that the box had stuck again and that the pump was not starting at all. By the time we had towed the car home it had of course reset and was all working again but this was finally the evidence we needed - the relay was working but the pump had stuck even though most of the time it was working perfectly.
I was now able to swap the pump out with the one from the spare robot. That was 3 months ago since when the car has driven faultlessly doing long and short trips with no stalling and no gearbox errors. If I hadn't already had the pump under the bench the cost would have been around £200 for a new one.
So it's true what they say about intermittent faults being the hardest to trace and I pity any owner or garage who is faced with a problem like this. Given that the software was not at all conclusive about any of the issues it's totally understandable that the only solution a dealer would offer is a replacement unit. But it goes to show that with persistence and probably more than a little luck problems can sometimes be solved.
I hope my story may help someone out there resolve a similar problem as I have been helped by others countless times in the past. Thanks all and keep up the great work on the FIAT forum!