Amazon Launches $250 Ring Drone and $1,000 Rover: An AI Patrol With 8 Minutes of Flight

An ordinary security camera looks at a single spot. If a burglar comes in from another side, you will see it at best at the edge of the frame. Amazon proposes a radical fix: let the camera drive or fly around the house on its own. On November 10, the company introduced the Ring Robotics lineup: the Ring Drone Cam drone and the Ring Rover Cam wheeled robot. Both run on the AI Pilot system.

How it works

AI Pilot combines the device's cameras, "specialized computer vision models," onboard sensors and cloud processing. First, the owner either walks the drone through the space by hand or lets the Rover explore on its own. The system builds a floor plan and recognizes rooms. It then suggests patrol points: by side doors and windows or, in a business, say, at the cash register.

From there, everything runs on triggers. When a Ring camera, sensor or alarm goes off, the device drives or flies along a preset route to see what happened. For the drone, this works with a Ring Protect subscription. It can be controlled from the Alexa Home Map, where all Amazon devices are collected.

On specs, the Ring Drone Cam starts at $250, the same as its predecessor, the 2021 Always Home Cam, which was sold by invitation only. It has a 148° field of view, 360° rotation, 2K video, two LED lights for the dark and a docking station. But it flies for only 8 minutes, and that is on a new battery and under controlled conditions. Sales are scheduled for January 27, 2027, but only if the FCC grants approval.

The Ring Rover Cam costs $1,000 and is positioned as a robot for business. Preorders are open, and delivery begins on December 16, 2026. It records video up to 4K, has a retractable mast for peering over counters and shelving, plus a siren, a light tower and a microphone for two-way communication. The stated coverage area is about 465 m². Recordings are stored in the cloud for up to 180 days.

What Amazon honestly states in the limitations

The most useful part for engineers and product managers is in the product description. The drone navigates worse on polished marble, wet tile, bare matte walls without texture, and near mirrors and frameless glass. The Rover cannot climb stairs and dislikes curved edges, recesses in the floor, thresholds higher than 3.5 cm and black glossy floors.

These are typical weak spots of visual navigation, which relies on distinctive points in the frame. A real home, such as a mirrored entryway with a glossy floor, turns out to be an edge case for it. If you are building a physical AI product, it is better to put such limitations in the documentation up front rather than leave users to discover them on their own.

The trust problem

The main risk here is not technical. In 2023, the FTC accused Ring of letting employees and contractors view recordings from customers' private spaces. Because of weak protection, hackers gained access to 55,000 accounts in the US. Now Amazon promises "privacy and security without compromise," LED recording indicators and video protection with Throw Away the Key encryption, which appeared in September 2026.

The EFF considers it an improvement but far from end-to-end encryption. Advocates also criticize the course set by Ring founder Jamie Siminoff and the return of features through which police can request recordings. Researcher Ryan Calo warned back with the first drone that a flying camera gives the company far more detail about the insides of homes and offices. Amazon has not yet announced facial recognition in its new robots. But since December 2025, Ring's stationary cameras have had a feature called Familiar Faces, and a class action over it is already under way.

What to do with this

For developers and product managers, the takeaway is simple. A mobile camera with AI is technically possible already, but reliability is limited by battery, floors and mirrors. You will pay for it with a subscription, and you will have to trust the manufacturer's reputation. If you are building similar products, ask yourself what exactly goes to the cloud, who can view it and how you will prove that to the user. For founders, it is also a reminder: a brand's privacy history never goes away; it travels with the brand to every new device.