HANDSWho we are
Enthusiasts who noticed a gap.
HandsFree started as an argument about whether Full Self-Driving was good enough to earn money with. The only way to settle it is to try, count everything, and publish the count — so that is what we are setting up to do. Nothing below has happened yet.
The gap, plainly
Driverless cars are coming to America city by city, biggest first. The Tri-Cities is not on anybody's list and won't be for years.
Tesla's robotaxi service runs in six cities across Texas and Florida. Waymo is rolling out in Nashville. Both companies expand by metro population, because that's the only way the economics work at their scale — mapping a city, staffing a depot and stationing a support team is enormously expensive per city.
Which means somewhere around 200,000 people here get the technology last, or get it from somebody local who was already doing it. We'd rather it was the second one.
Meanwhile Tennessee has, as we read it, permitted vehicles without a human driver since 2017 and bars cities and counties from banning them outright. If that reading is right, the legal door has been open here for years and nobody local has walked through it. We are having it checked properly rather than relying on our own reading.
- Market
- Johnson City, Kingsport, Bristol — roughly 200k people, one airport, one large university
- AV competition
- None that we can find. Nearest announced is Nashville
- State law
- Driverless permitted since 2017; local bans prohibited
- Rideshare supply
- Our impression is that it thins out late and around ETSU term dates. Worth checking properly, not a study
- Airport
- TRI. Steady, schedulable, high-value, and the runs are long enough to be worth doing
- Runway before a giant arrives
- Our estimate: 3–6 years. A guess, not a measurement
What we believe
Four things, and we'd rather be held to them
01
Show the working or don't make the claim
Anyone can say their service is safe. California makes companies testing autonomous vehicles file an annual public disengagement report, and it is a genuinely good dataset — but it is a legal requirement, it is annual, and it covers testing rather than carrying passengers. No rideshare service publishes weekly, unprompted, in a state that asks for nothing. That's not a hard bar to clear, and clearing it is most of our strategy.
02
Tell people before they find out
Every passenger will be told the car is running FSD before the wheels move. Every operator will be told the job has an end date before they sign anything. Both of those cost us something, and both are cheaper than the alternative.
03
Being small is the advantage, not the handicap
A local fleet ends up knowing which junction confuses the car, because it drives the same roads all day. A national fleet can't hold that. Depth in one small place beats thin coverage everywhere, right up until it doesn't — and that's years away.
04
The interesting version is the marketable version
We're not going to out-spend Uber. We can be the thing people film and send to their brother. A screen showing what the car sees and a sign that winks at the traffic come to about $180 a car, and we are betting they beat any ad we could buy at that price.
- That riders will be comfortable with it. We think most people will shrug and a few will ask for it off. If it turns out to be most people asking for it off, the whole proposition is weaker than we think and we will say so here.
- That winter is manageable. Cameras struggle in spray, salt and covered markings. We have planned for a worse winter intervention rate and shifts driven manually, but we have not lived through one as a fleet.
- That publishing the data earns anything. It might turn out that nobody reads it and it is pure cost. We are doing it anyway, because the version of this company that quietly stops publishing when the numbers get awkward is not one worth building.
- That the runway is years, not months. If a large operator announces the Tri-Cities sooner than we expect, our head start is smaller than the plan assumes.
This list is deliberately written now. Once we are running, the things we actually got wrong replace these — and the originals stay up.
Track record
The mistakes will stay up.
A company that only publishes its wins is publishing advertising. The list beside this is permanent, it grows, and nothing ever comes off it.
There's a straightforward selfish reason for it, too. In a few years we'll need somebody to write an insurance policy for driverless operation on our specific roads. When that conversation happens we want to arrive holding a long, boring, unedited record — including the parts where we got it wrong and said so at the time.
You cannot manufacture that retrospectively. You can only have started early.
handsfreefsd.app
Come and have a look
Ask to be told when we launch, put your Tesla to work, or read how we have promised to measure ourselves and tell us where it's wrong.