Self-driving golf carts: How dealers and OEMs can profit right now

Yes, robotaxis proved people will pay for a driverless ride, but autonomous golf carts represent a business that dealers and OEMs can run right now.

At last week’s 2026 GolfCarting Expo & Dealer Summit in Charleston, Joyride Founder and CEO Vince Cifani gave a talk called “Autonomous Advantage” to a room full of dealers and manufacturers who are all wondering the same thing: with robotaxis now carrying hundreds of thousands of paying riders every week, what does autonomy mean for the people who build, sell and service golf carts?

The answer is that the golf cart industry is better positioned for self-driving vehicles than the car industry. In this post, we walk through the data: where autonomous vehicles stand today, what riders pay and why courses, campgrounds, resorts and communities are the places where low-speed autonomy makes the most sense.

Key takeaways

• Autonomy works today where the environment is contained. Golf courses, campgrounds and resorts are among the most constrained environments there are.
• Riders pay a 12% to 27% premium for driverless rides in San Francisco, and 70% of people who have tried a Waymo prefer it to a human-driven ride.
• A self-driving vehicle is built from five parts, and four of them (sensors, compute, autonomy software and the fleet layer) are now available off the shelf.
• Some of the most valuable companies in the robotaxi industry build no autonomy technology at all. They operate, charge and service the vehicles, which is work dealers already do.

What is a self-driving golf cart?

A self-driving golf cart is an electric cart that can steer, accelerate, brake and navigate on its own within a defined area, such as a golf course or resort, without a person at the wheel. In industry terms, it is a Level 4 autonomous vehicle.

The SAE scale runs from Level 0 (no automation) to Level 5 (a vehicle that can drive anywhere a person can). Almost everything on the market today sits at one of two levels:

• Level 2, partial automation: the system steers and controls speed, but a human must supervise at all times. Tesla FSD (Supervised) is the best-known example.
• Level 4, high automation: no human is needed, but only inside a defined operating design domain (ODD), meaning the specific area and conditions the system is built to handle. Waymo and Zoox robotaxis are Level 4.

Level 5 does not exist commercially. The step that matters is from Level 2 to Level 4, where the vehicle, not the person, is responsible for driving. Because Level 4 is defined by its operating domain, the smaller and more predictable that domain is, the easier autonomy becomes. A golf course is a domain you can draw on a single sheet of paper.

Where autonomous vehicles stand in 2026

Autonomy is no longer a research project. As of fall 2026:

• Waymo runs roughly 4,000 vehicles across 11 U.S. cities and delivers around 500,000 paid rides per week, double its volume of about a year earlier.
• Zoox, owned by Amazon, began charging for rides in Las Vegas on August 10 and received a federal exemption to operate up to 2,500 robotaxis without steering wheels.
• Tesla carried its first public Cybercab riders in Austin on September 3, in a vehicle with no steering wheel or pedals.
• Uber has committed $10 billion to deploy 120,000 driverless vehicles over the coming years, working with partners that include Nuro, Wayve, WeRide and Pony.ai.

Regulators are moving too. In June, the U.S. Department of Transportation proposed dropping the brake-pedal requirement for vehicles designed to drive themselves, and in August, Nevada approved up to 8,000 robotaxis in Clark County.

The most useful pattern is in the companies making money quietly, away from city streets. Gatik has $600 million in contracted revenue for driverless box trucks running fixed routes for customers like PepsiCo and Walmart. Bedrock Robotics runs fully autonomous excavators at customer sites. Beep has operated 15-mph autonomous shuttles in Lake Nona, Florida, a master-planned community, since 2019, and has raised about $130 million, most recently in a Series B led by Mobileye. Autonomy is furthest along where the environment is most controlled.

Do people pay more for a driverless ride?

Yes. A study by rideshare analytics firm Obi compared median fares on the same routes and times in San Francisco:

Waymo launched in San Francisco at a 30 to 40% premium over Uber and Lyft. That gap has narrowed to 12 to 27%, mainly because Waymo trimmed prices while Uber and Lyft raised theirs, but riders still pay more to ride without a driver. Obi’s survey also found that 70% of people who have taken a Waymo prefer it to a human-driven ride. Waymo now sells a $30 per month Premier subscription for priority pickup and cash back.

To be sure, the demand is not universal. About 40% of survey respondents said they would not pay more for a driverless ride, and San Francisco is an affluent, tech-forward market. Still, the lesson for anyone running a fleet is clear: driverless rides are being sold as a premium product.

The growth runway is long. Waymo holds roughly a 25% share of ride-hail in San Francisco, its most mature market, but only about 0.65% of the roughly 4 billion ride-hail trips taken in the U.S. each year. Waymo is importing around 300 new vehicles a month and is limited by how many vehicles it can put on the road, not by rider demand.

What is still hard about robotaxis on public roads?

The technology works, but running it on open roads is expensive and complicated:

• Edge cases: Waymo recalled about 4,000 robotaxis after vehicles drove into highway construction zones, had separate problems with flooded roads, and paused freeway service nationwide for more than two months.
• Cities and first responders: Waymos stalled and blocked streets in San Francisco on July 4, prompting the mayor to call for tougher rules. Minneapolis is considering requiring a person in the seat.
• Empty miles: 44% of Waymo’s California miles are driven with no passenger, according to UC Berkeley analysis of state data.
• Uneven growth: Tesla’s paid robotaxi miles fell about 36% quarter over quarter, and Cybercab volume production has slipped past 2026.
• Hardware cost: Waymo’s sixth-generation system uses 13 cameras, 4 lidar units and 6 radar units, and costs an estimated $20,000 to $25,000 per car. Its new vehicle is built in China and faces a 100% tariff.

Each of these problems comes from operating in an open, shared environment with strangers, traffic, weather, city politics and constantly changing streets.

Which leads to our next point…

Why are golf carts the right vehicle for autonomy?

A self-driving golf cart uses the same five-part technology stack as a robotaxi: sense, locate, think, act and connect. The difference is the environment it operates in.

The cost difference follows from this. A robotaxi company pays for a $40,000 to $100,000+ vehicle with an automotive-grade sensor suite, depots and remote-assistance staff in every city, permits market by market, and marketing to win riders from Uber and Lyft. A cart operator already owns the vehicle, the barn and the chargers, and serves guests who are already on the property and already renting carts. Trips are short and predictable, so a cart is either working or parked, rarely driving around empty looking for a rider.

Riders pay the same premium for a driverless ride. The cost of delivering it on private property is a fraction of the cost on city streets.

Where do self-driving golf carts make sense?

Golf courses, where carts that return themselves to the barn, reposition between holes and the range, deliver food, drinks and merchandise to players, and run marshal or pace-of-play routes.

Campgrounds and RV resorts, where guests hail a ride from their site to the pool, store or lake; carts deliver luggage and firewood; shuttles run after hours without a staffed driver.

Resorts, communities and venues, where on-demand shuttle loops in master-planned communities, hotel-to-beach runs, stadium and fairground parking lots, and college campuses.

Industrial and institutional sites, where people movers and tow vehicles on fixed routes at data centers, plants, hospitals and airports.

Beep’s shuttles in Lake Nona show this model already works at low speed. Beep also started out as a vehicle dealer in Florida before becoming an operator and software company, which is a path many golf cart dealers will recognize.

How can golf cart dealers and OEMs profit from autonomy?

Most of the autonomy stack can now be bought rather than built. Sensors such as short-range solid-state lidar and cameras have fallen in price thanks to automotive volumes. Embedded AI computers are available off the shelf. Autonomy software has become something companies can license: in August 2026, Nvidia released Alpamayo 2 Super, a driving model that can be used commercially. The fleet layer, covering dispatch, telematics, keyless access, payments and remote assistance, is already mature and in use today for golf cart rentals.

The scarce part is a vehicle that software can drive. That is where dealers and OEMs come in. The robotaxi industry offers three business models that translate directly:

1. Dealer-led: retrofit existing fleets

Add an autonomy kit to electric carts already running on a customer’s property. This is the fastest route, and it lets a dealer prove utilization and willingness to pay before anyone buys a new vehicle. Waymo followed the same path, starting with adapted Jaguar vehicles before moving to purpose-built ones.

2. OEM-led: sell autonomy-ready vehicles

Build drive-by-wire steering, braking and throttle, plus sensor mounts, a compute bay and power, into carts at the factory. Every unit sold can be upgraded to autonomous later. Stellantis is already marketing “L4-ready” vans to autonomy companies, and a golf cart OEM can do the same with far less engineering work. This route carries the biggest margin.

3. Dealer plus partner: run autonomy as a service

A software platform like Joyride provides the autonomy, while the dealer runs the fleet, charging, service and customer relationship, and bills per ride or per month. This is the model behind Moove, which raised $250 million at a $2.1 billion valuation as a fleet operator that builds no autonomy technology. It now runs Waymo fleets in Phoenix, Miami and Las Vegas. Tesla has started recruiting fleet buyers for the Cybercab as well.

These models can be combined. A dealer can retrofit carts today, push its OEM for autonomy-ready platforms in the next model year, and build a service business on top of both.

New revenue lines for dealers

• Per-ride and per-hour pricing: moving from flat daily rentals to usage pricing, with riders in cities already accepting a 12 to 27% driverless premium.
• Subscriptions: a resort or community can sell a seasonal ride pass, the same way Waymo sells Premier for $30 a month.
• Labor savings for customers: cart retrieval, repositioning and off-hours shuttle driving are the first jobs autonomy replaces, which becomes the dealer’s sales pitch.
• Service and calibration contracts: sensors and computers need cleaning, calibration and firmware updates, creating recurring service revenue that OEMs cannot easily take direct.
• Software and data: fleet software, utilization reports and route analytics bring in recurring revenue.
• Higher-value vehicle sales: an autonomy-ready cart carries a higher price and a longer customer relationship, the same way connected and keyless carts became worth more than standard ones.

What to watch over the next 12 months

Here are signs the timeline is speeding up:

• The federal rule removing steering wheel and pedal requirements is finalized.
• Tesla sells Cybercabs to third-party fleet operators, creating a template for OEM-to-dealer autonomous vehicle sales.
• Small vehicle makers adopt open driving models like Nvidia’s Alpamayo.
• Low-speed autonomous pilots appear at resorts, campuses and communities.

Here are risks to watch:

• State or city rules requiring a person in the seat spreading to LSV regulations.
• Insurance markets tightening on autonomous coverage before enough data exists.
• New tariffs or security reviews on Chinese-made sensors and vehicle parts. Idaho National Laboratory is already reviewing Chinese lidar.
• A high-profile low-speed incident anywhere shifting public opinion.
• Vendors promising “Level 5.” Always ask what the operating domain is.

Why now is the time for self-driving golf carts

In 2026, riders showed that they will pay more for a driverless ride. The hardware and software needed to build one became affordable and available. And the robotaxi industry showed that the hard part of autonomy on public roads, operating in an open and unpredictable environment, is the part golf carts do not have to deal with.

The properties dealers already serve are some of the best places in the world to run autonomous vehicles. Even the industry leader has reached less than 1% of U.S. ride-hail trips, and the resort, course and campground market has barely been touched.

Joyride’s platform already makes more than 10,000 electric vehicles in over 550 markets keyless, connected, shareable and trackable, which covers the fleet layer of the autonomy stack. We have been working on what comes next for the dealers and OEMs we work with, and we will have more to share soon. In the meantime, if you want to talk about what an autonomy pilot could look like for your fleet or your customers, reach out to our team today.

Frequently asked questions

Are self-driving golf carts legal?

On private property such as golf courses, campgrounds and resorts, operation is mostly governed by the property owner’s rules, along with low-speed vehicle (LSV) regulations where they apply. This is a much simpler situation than robotaxis face on public roads, where federal exemptions, state agencies and city councils are all involved. Rules vary by state, so check local LSV regulations and talk to your insurer before launching.

How do self-driving golf carts work?

They use the same five functions as a robotaxi: sensors (cameras, lidar, ultrasonic) to see the surroundings, GPS and maps to know their exact location, an onboard computer to plan a path, drive-by-wire controls to steer, brake and accelerate, and a connection to a fleet platform for dispatch and remote assistance.

Do customers pay more for driverless rides?

In San Francisco, riders pay a 12 to 27% premium for Waymo over Uber and Lyft, and 70% of people who have tried Waymo prefer it to a human-driven ride. About 40% of riders say they would not pay more, so pricing should be tested during a pilot.

Can existing golf carts be made autonomous?

Yes. Electric carts can be retrofitted with drive-by-wire controls, sensors and an onboard computer. Retrofitting carts already on a customer’s property is the fastest way to test autonomy before buying new autonomy-ready vehicles.

What is the best first use case for an autonomous golf cart?

Start with one repeated task on one property, such as carts returning themselves to the barn or a fixed shuttle loop at a campground or resort. Short, predictable routes make it easier to measure results and keep the pilot safe.

How can golf cart dealers make money from autonomy?

Dealers can earn from per-ride and subscription pricing, recurring service and calibration contracts, fleet software and reporting, higher-value vehicle sales, and the labor savings they deliver to property owners. Dealers can retrofit existing fleets, sell autonomy-ready carts from OEMs, or run autonomy as a service with a software partner.

Data in this article comes from the “Autonomous Advantage” presentation at the 2026 GolfCarting Expo & Dealer Summit, drawing on TechCrunch Mobility, Driverless Digest, Obi rideshare pricing studies, IIHS, CPUC filings and UC Berkeley analysis.