Right foot bruise, the driver must be from the UK. That is a serious bruise. I would document everything. Lots of pictures, videos, of the other car, etc. Report it to your insurance. Report it to whatever automotive safety oversight agency there is in the UK, that might be able to come look over the scene while it's still fresh.
Location says Greenmount, Maryland. So, it's the U.S..
To the OP: Picture of the car/your daughter's car, or it didn't happen.
Brand new car, you're new to the forum, that's OK.
You'll run into lot of flack; that might put you off. Don't worry about it. Several things, in no particular order:
- There's safety things up the wazoo that, if this did happen, had to have failed in order for it to happen.
- I diagnose large, complex, software/hardware driven electronics for a living. There's three things:
- The Bathtub Curve. X axis: Time, sometimes log time. Y axis: Probability of failure. At t=0, with spanking new hardware, there's a relatively high failure rate because Brand New Stuff sometimes comes with manufacturing defects. Those die early on. After one gets past this, there's Poisson-distributed random failures, at a low level, over time. Finally, at the End of Life, we get wear-out errors, and the probability of failure goes up. The shape looks like a cross-section of a bathtub.
- Infant Mortality. Failure rates are high at the beginning of life of any product. It's why test equipment, when built, gets (typically) baked at 100F for hours to days before being sent to customers, the idea being to weed out the weak ones. They don't bake entire cars; but failure rates of components are measured. Those with high failure rates get redesigned, parts, if necessary, get extra inspections, and Screaming At Vendors when bad lots of parts show up is a Thing. This is the reason that cars (and pretty much everything else) has warranties.
- FIT rate. Stands for Failures in 10^9 hours. Handles the Poisson-distributed stuff. A single resistor has a FIT of 1.0. Complicated ICs can have FIT of 10-100, depending upon I/O. (Interestingly, the more I/O on a device, the higher the FIT tends to be.). Add up all the FITs of all the parts, and, typically, the total FIT is the FIT of the board. A typical PC has a FIT of around 2000-3000. So: If one had a board with billion resistors on it, one would expect one failure, per hour, on average. But if one has boards with 1000 resistors on them (that's easy), and one has built 1,000,000 boards, then one would expect one board to fail an hour. This is why auto manufacturers have service locations.
- Having said all that: The question now is, if your story is true, what are the minimum number of things that must have failed to allow the car to do what you relate? We're likely talking sensors, here, since those are (typically) physical parts that can die/snap. Door switch? Seat belt switch? Device that picks up these things at some central module and provides that information to the car computer (bad bus with all 0's coming out..) A lot of such mixed SW/HW problems are amenable to diagnostic tests, not to mention parity and CRC checks to detect stuck-at faults and such.
- Having said all the above: None of this kind of error checking is new or limited to Tesla. Car manufacturers have had fail-safe on drive-by-wire systems; gas pedal angle, electronics tests, etc., for a couple of decades now. But.. stuff happens.
First things first. As I said, first we need more than a picture of your bruised foot. I used to hang out on Tesla, Inc.'s forums before they got shut down. They got shut down in large part due to trolls, short-sellers, and other malefactors seeking to harm Tesla in one form or another, and Tesla's unwillingness to dedicate the necessary expense to moderate those forums. One of the Bad Guys tricks was a one-and-out: Post some utterly false, damaging statement, and then never show up again.
On the other hand, sometimes we
did get people with odd-ball stuff happening. Back in 2018/2019, M3's were new, there were a lot of new users, stuff happened: Busted roof/rear window glass, etc. Those people, we helped and gave solid advice to. I'd say about 3/4 of the mad-as-hell posters who showed up had, well, user errors. It ain't an ICE. The other 1/4 had varying problems, usually solved with warranty repair.
So, ante up. Got some pictures of the damaged cars? How about a VIN number, or at least when the car was manufactured? What was Tesla's reaction when you told them? We're willing to talk, but you got to talk back.