RIVR TWO is a wheeled-legged delivery robot powered by Physical AI that navigates real streets and completes unattended doorstep delivery without human handoff.
Photo source:
Rivr Two
Delivery networks have become remarkably
efficient at moving parcels and food across cities, but the final stretch, from
the vehicle to the customer's door, remains one of the most operationally
complex and costly parts of the entire journey. Sidewalks, curbs, steps,
crossings, and narrow entryways all present obstacles that wheeled vehicles
cannot handle and that human couriers navigate instinctively. Solving the last
100 yards of delivery without relying entirely on human labor has been one of
logistics' most persistent unsolved problems.
RIVR TWO
is a wheeled-legged delivery robot built by RIVR that combines physical
mobility with Physical AI to navigate real-world urban environments and
complete doorstep deliveries autonomously. Unlike conventional delivery robots
that operate only on flat, predictable surfaces, RIVR TWO is designed to handle
the full range of obstacles that define real streets, including curbs,
crossings, steps, and tight entryways, going beyond the last mile to solve what
RIVR calls the last 100 yards. It is built for both food and parcel delivery,
optimized for mixed fleet operations, and designed to integrate with existing
delivery vans without overhauling current operations.
The most distinctive engineering decision in
RIVR TWO is its purpose-built payload system. The robot is loaded from the top
and releases deliveries from the bottom by sitting down at the doorstep,
allowing parcels and food to be left without any direct human handoff or
customer interaction required. This top-load, bottom-release mechanism is not
simply a convenience feature. It is a fundamental rethinking of how robotic
delivery should work, eliminating the doorstep dwell time that slows down
conventional delivery workflows and enabling fully unattended operations at
scale. The payload capacity stands at 55 liters, covering the full range of
typical non-bulky parcel and grocery delivery needs.
What enables RIVR TWO to navigate environments
that have historically defeated delivery robots is its Physical AI system,
built on a combination of simulation data and real-world operational
experience. The robot is designed to navigate safely around pedestrians,
cyclists, and vehicles, with instant stop capability available through its
autonomy system, a physical button, or a remote command. It handles sidewalks,
bike lanes, crossings, and long distances without getting stuck, travels faster
than walking pace to support peak-time delivery demands, and uses lights and
visual cues to maintain visibility in both day and night conditions.
RIVR TWO is deliberately designed with a small
footprint and a soft, approachable form that allows it to operate in shared
public spaces without creating friction with pedestrians or communities. Its
design reflects an understanding that a delivery robot operating in urban
environments must feel like it belongs there, not like an intrusion. The robot
supports quick turnaround charging aligned with delivery shift patterns, making
it operationally viable for the continuous demands of real delivery networks.
Please subscribe to have unlimited access to our innovations.