Resources
Minneapolis Fought Over the Driver’s Seat and Ignored the Curb
Minneapolis spent an hour of debate and a veto on whether a person must sit in a robotaxi. The state will probably decide that. The curb, which the city controls and has already measured, got no rule.
By Greg Bessoni, Founder and CEO, XoomPark ·
On October 1, 2026, the Minneapolis City Council passed the Workers Drive Minneapolis ordinance by a vote of 7 to 6, after more than an hour of debate.1 On October 2, Mayor Jacob Frey vetoed it.2 “Requiring a driver in a driverless car doesn’t make sense,” he said.3 An override takes nine votes, and the ordinance had seven.4
What the ordinance actually required
Most coverage said the ordinance put a person in the car. It did more than that, and the details are where the argument lives. It created a city-issued license for any company operating autonomous rideshare in Minneapolis.5 It required a safety monitor in the driver’s seat during operation, with “immediate access to steering, braking, and acceleration.” The monitor had to hold a valid Minnesota driver’s license and be paid at least minimum wage.3 None of it took effect until October 2027.
Two of those details matter more than the headline. The pay floor makes the monitor a job, not a safety feature, and the sponsors did not pretend otherwise. The 2027 date means the council wrote a rule for a service that does not yet exist in the city, with a year left for the state to act first. Both details tell you what the fight was about. It was about who gets paid to be in the car.
The strongest case on each side
The sponsors have a constituency, and it is not small. Council Vice President Jamal Osman co-authored the ordinance and frames it as worker protection: if the technology takes jobs out of Minneapolis, it is a bad deal for Minneapolis.6 The sponsors’ joint statement warns of a “10,000 driver workforce shrinking.”2 Ten thousand people in one city is not a rounding error, and a council that ignored them would not be doing its job. Robin Wonsley, Jason Chavez, Aurin Chowdhury and Osman are urging an override.7 They need two more votes. No override vote has been reported or scheduled as of October 12. The ordinance is vetoed, not dead.
Frey has a real point too. A rule that puts a person in the driver’s seat regulates the vehicle, not what the vehicle does to the street. It says nothing about where a car may stop, how long it may sit, or who answers the phone when it blocks a lane. “Thoughtful regulation for safety is necessary, but not through a backdoor ban,” he said.4 His office had already called the measure “a back-door ban, and a bait for the state to come in and preempt us.”8 Waymo called it “a de facto ban on autonomous vehicles.”4
We are not going to pick a winner on the labor question. It is a real question, and it is not ours to settle. What we can say is that both sides spent the whole fight on the one lever the city is least likely to keep, and neither side touched the lever it will certainly keep.
The fight was over the lever the state will take
A driver mandate is vehicle regulation. Who may operate a vehicle, with what equipment, under what license, is the kind of rule states write, and when states write it, city rules on the same subject usually fall. Minnesota has no broad autonomous vehicle law. Its 2026 session ended with no bill passed, even though both parties said rules were needed.9 No 2027 preemption bill has been reported, so the timing is unknown. The shape of the risk is not. If the state writes a vehicle rule, the city’s vehicle rule goes with it. Frey’s office said as much with the word “bait.”
The curb is a different kind of power. Which stretch of curb is a loading zone, a bus stop or a pickup zone is a local right-of-way decision. How long a vehicle may sit there is a parking rule. Where a fleet holds its cars between trips, on a lot or in a garage instead of a travel lane, is land use. Those powers stay with the city when the state takes the driver’s seat, and they are the powers that decide what a resident actually experiences: whether the car waits in a bike lane outside her building or on a lot two blocks away.
Bern Grush of the Urban Robotics Foundation wrote in June that early robotaxi cities have seen vehicles “blocking bus stops, idling in bike lanes, and impeding emergency vehicle access.”10 He calls these “curb management failures” and says they “are preventable,” and he argued the city could set pickup and drop-off rules before launch. The council did not argue about the curb. The veto did not propose anything for it either.
Minneapolis is unusually well placed to act. Under a federal SMART grant it built a digital inventory of its curb space, a digital twin of about a mile of Nicollet Avenue, and the data showed compliance with existing curb rules “lagging.”11 In plain terms, the city measured the problem before any robotaxi carried a passenger. It has the map. It has not written the rule.
What a curb-first ordinance would say
These are performance standards. They describe outcomes the city can measure, not a vendor or a product.
- Staging off the street. A licensed operator holds vehicles between trips off the public right-of-way, with capacity the city sets as a share of vehicles in service, and may not use curb or travel lanes as a waiting area.
- Defined pickup and drop-off zones. Rides begin and end at zones the city designates, not wherever an app drops a pin. The operator’s routing must send the car to the nearest zone, and the city publishes the zones as open data so any operator can comply.
- Dwell limits with enforcement. A maximum number of minutes at any zone, reported by the operator per zone per day, audited against the city’s own curb sensors, with a fee schedule that rises with repeat violations.
- Entry and exit data from the operator. Trips, dwell time and staging locations by zone, on a schedule and in a format the city sets. The SMART grant work already built the shelf this data sits on.
- Incident recovery with a responsible party. A named contact reachable around the clock, a response-time standard for a stopped or stuck vehicle, and a penalty when the standard is missed.
Every item above survives a state vehicle law, because none of them says who sits in the seat.
The window
Waymo has tested in the Twin Cities with people behind the wheel since November 2025 and has not carried a paying customer in Minneapolis.3 The city holds curb data nobody has acted on. In our reading, the cities that waited until after launch wrote their curb rules under more pressure and got worse ones.
Sources
- Star Tribune: Minneapolis City Council passes divisive ordinance on driverless vehicles, 7-6 (Oct. 1, 2026)
- KSTP: Mayor Frey vetoes the robotaxi ordinance, with the sponsors’ statement (Oct. 2, 2026)
- MPR News: Mayor Frey to veto Minneapolis council-approved robotaxi regulation (Oct. 1, 2026)
- CBS Minnesota: Minneapolis City Council approves driverless rideshares ordinance; Mayor Frey says he’ll veto (Oct. 1, 2026)
- FOX 9: Minneapolis Mayor on robotaxi regulation veto: ‘Doesn’t make sense’ (Oct. 2, 2026)
- KARE 11: Mayor Frey vetoes city council’s Waymo regulation (Oct. 2026)
- Patch: Frey vetoes robotaxi rules; council members push override (Oct. 7, 2026)
- Minnesota Reformer: Minneapolis weighs ban on driverless robotaxis to protect rideshare drivers (Sept. 17, 2026)
- CBS Minnesota: Waymo’s future in Twin Cities uncertain after Minnesota Legislature fails to pass guidelines for self-driving cars (May 20, 2026)
- Urban Robotics Foundation: Can Minneapolis Get Ahead of the Curb? (Bern Grush, June 14, 2026)
- Government Technology: Minneapolis, Seattle partner on digital curb management (Jan. 7, 2025)
Sources are public. XoomPark isn't affiliated with the companies referenced in this post.
About the author
Greg Bessoni is the founder and ceo, xoompark. He writes about fleet operations, curb logistics and AV infrastructure.