Notes for Zoox's warehouse and logistics team
Today's whole fleet, built in One week
Zoox has about 100 robotaxis on the road and has said Hayward can ramp to 100 a week. From there, the plant can build faster than cities open, so the hard part becomes matching what's built with when each city launches: finished robotaxis waiting to deploy, the space to stage them, and the parts behind them. Here is how I see it, and how I'd help.
Built from public information only. Assumptions are labelled.
Robotaxis a week.
Opening: ~1 a day (CNBC, June 2025). Exemption: NHTSA allows up to 2,500 commercial vehicles a year (July 2026), shown spread over 52 weeks. Hayward ramp: the 100 a week Zoox has said the plant can reach, subject to regulatory approval. Full scale: 10,000 a year.
I'm Tristan, a 2nd-year Industrial Engineering student at the University of Toronto. At Technimount, a manufacturer in Québec, I built an MRP system that cut order processing time by 90% and automated procurement from quote to purchase order, and I now build the production board that tracks inventory along their assembly line.
The analysis
Tap to open. Built from public information; assumptions labelled.
1. One week of output
At 100 a week, the plant outruns the exemption in 25 weeks.
Full scale (10,000 a year) is 4× what the exemption allows. So from here the build rate follows the launch calendar, not the other way around. Unless the rate flexes, every week a launch slips at 100 a week is up to 100 more finished robotaxis to park.
2. Build vs deploy
Launches are lumpy and the line is level, so robotaxis wait even when nothing slips.
My example: eight cities from Zoox's public list (Las Vegas, San Francisco, Austin, Miami, Dallas, Phoenix, Los Angeles on Uber by mid-2027, Atlanta), with launch weeks and fleet sizes I made up, 2,150 robotaxis in all, built at a level 60 a week, two weeks from build to launch. The numbers move with those guesses; the pattern doesn't. That trade-off is what an S&OP process decides: flex the rate, or plan the parking.
3. Where it's decided
Inbound material, finished-robotaxi staging, and service parts.
The flow as Zoox describes it publicly. Highlighted steps are my hypotheses.
- WatchInbound material and component storageSuppliers preassemble major parts; Hayward stores them. At 100 a week, every extra day of material on hand is 20 more robotaxi sets on the floor (5-day week). The new 350,000 sq ft next to the plant is there to scale production.
- Assembly lineAGVs move each robotaxi station to station; about 20 minutes of work per vehicle at the line's pace, so 100 a week is about one full shift (my reading).
- End-of-line testSensor calibration, alignment, dyno, water test, light tunnel, test track.
- WatchFinished robotaxis waiting to deployBuilt, tested, not yet launched. Zoox already has a 34,000 sq ft logistics and staging site in Hayward, and Santa Monica is negotiating space for up to 150 robotaxis. How many can wait, and where?
- Ship and launchCity by city, each with its own permits.
- WatchService parts for the fleetThe redesigned robotaxi joins the fleet as it comes off the line, next to the ~100 already running, so parts plans cover two versions and every cut-over (my inference).
4. A model you can run on your numbers
Launch calendar, build rate, robotaxis waiting, and the line behind each rate.
- Build vs DeployPut in the real launch weeks and fleet sizes: it gives the lowest level rate that meets them, how many robotaxis wait each week, and what a slip adds.
- Line and MaterialLine hours, shifts and material on hand at each rate, from opening to full scale.
- InputsThe public numbers, with sources: change any of them and everything recalculates.
Blue cells are inputs; yellow cells are my assumptions, to replace with real data.
My 16 weeks
- Wk 1–2Learn the flow. Receiving, storage, line-side delivery, finished-robotaxi staging and shipping, on the floor with the people who run them; map one upcoming city launch end to end.
- Wk 3–6One weekly view of build vs deploy. Built, tested, staged, shipped and launched, by city, with staging space and days of material on hand next to it.
- Wk 7–11Test it against the next launches. Where staging, inventory accuracy or service parts run short first, with Manufacturing, Fleet Operations and Procurement.
- Wk 12–16Run it and prove it. Support one launch with it, measure how long robotaxis wait before and after, and leave it ready for the next cities.