As drone delivery takes flight in the suburbs, a different kind of compact, driverless vehicle is making inroads in cities.
Ground-based robot delivery companies are transforming the urban final mile by using every part of the road to make deliveries. Sidewalks, bike lanes, margins, you name it — autonomous robots from Starship Technologies, Nuro, Kiwibots and others are smart enough to take deliveries anywhere. And they’re being deployed by brands like Safeway, Dominos and merchants using delivery platforms such as Uber Eats (NYSE: UBER) and FedEx (NYSE: FDX).
Add Chick-fil-A to the list. The maker of one of America’s most popular chicken sandwiches is partnering with Refraction AI, one of the newest players in the growing robot delivery space, to deliver from two Chick-fil-A restaurants in downtown Austin, Texas. The commercial pilot will begin late in June.
“We are thrilled about working with Chick-fil-A, an organization that is admired and respected as much for its commitment to the communities it serves, as it is for the innovation and quality of its business,” said Refraction AI CEO Luke Schneider.
It’s unclear just how many robots Chick-fil-A will receive in this initial launch. But Refraction AI co-founder and CTO Matthew Johnson-Roberson said that it would be “”an order of 10,” adding that the deal is part of a larger program to test the efficiency, profitability and sustainability of robot delivery for quick-service restaurants.
“Refraction AI was founded as a way to unlock the many benefits of affordable, sustainable delivery using autonomy technology that is available today. Scaling the last-mile — making it work for our communities, restaurants, grocers and retailers — is a key step in the journey to a safer, cleaner, brighter future,” Johnson-Roberson said in a statement.
Like many other delivery robots, the company’s flagship REF-1 robot can switch between full autonomy and remote operation when needed. It also has an insulated compartment to maintain the temperature of the food, but the company aims to deliver within 10 to 12 minutes to ensure orders are received at the proper temperature.
The REF-1 is unique among delivery robots in a few ways, however. For one, it’s larger and heavier than sidewalk robots from Starship Technologies or Kiwibots but smaller and lighter than a car-size robot like Nuro’s.
It’s also lightweight and low-power enough to qualify under e-bike regulations but fast and nimble enough to keep up with vehicles on the street. That makes it something of a Goldilocks, capable of operating in both bike and car lanes without impeding traffic.
On sidewalks or shoulders, the REF-1 typically travels at around 15 mph. That’s slow enough to give it a stopping distance of about 10 feet and allows Refraction AI to use low-cost sensors like cameras rather than expensive lidar, which is needed for the vehicle to “see” farther away. Robots like Nuro’s, which travel at higher speeds, require more costly technology.
Watch: More Providers Offering Robotics as a Service
But perhaps the most intriguing difference between Refraction AI and other robot delivery players is the platform itself. It’s one of the first companies to deploy a robots-as-a-service (RaaS) model — essentially a subscription to robots — for delivery. Under an RaaS framework, brands can sign a contract to use a fleet of delivery robots for a set time period, typically one to three years, without buying the vehicles outright.
The model is one that has seen rising adoption in the warehouse automation space. In the warehouse, operators are leveraging short-term contracts to buy automation on the cheap rather than fronting cash to buy an expensive fleet of robots.
According to the company, Refraction AI is looking to bring RaaS outside of food delivery. It’s inking partnerships with retail and grocery partners, a spokesperson told TechCrunch. Parcel delivery is another potential avenue for the platform — Uber launched same-day parcel delivery in 6,000 U.S. cities using its fleet of contracted gig workers, and Refraction AI could do something similar with its fleet of rentable delivery robots.