Delivery robots are becoming a common sight in some cities, but they've had one major limitation – they still need humans to load and unload them. A clever new robot called LEVA is changing that by picking up its own cargo, navigating through city streets, and even climbing stairs all on its own.

Not Just Another Delivery Bot

Developed by researchers from ETH Zurich, Zurich University of Applied Sciences, and Zurich University of the Arts in Switzerland, LEVA combines the best features of wheeled vehicles and walking robots. The result is a versatile machine that can handle tasks that would stump most delivery robots.

"We wanted to create a robot that could navigate urban environments completely independently, from picking up packages to final delivery," explains Dr. Lisa Müller, one of the project leaders. "No human assistance required at any step."

What makes LEVA special is its ability to spot standard Euro-type cargo boxes, position itself correctly, and then lift boxes weighing up to 85 kilograms (187 pounds) – about the weight of a large adult.

Rolling and Walking: The Best of Both Worlds

LEVA looks a bit like a table on wheels, but there's much more to it. The robot has four legs with motorized wheels at the ends instead of feet. This hybrid design gives it unique capabilities:

- On smooth surfaces like sidewalks, it rolls like a car, which is faster and uses less battery power
- When it encounters stairs, curbs, or rough terrain, it can lock its wheels and walk like a four-legged animal
- Its wheels can turn independently, allowing it to move sideways or in any direction without turning around

This versatility means LEVA can navigate complex urban environments that would stop typical wheeled robots in their tracks.

How It Picks Up Boxes

The box-lifting process is fascinating to watch. LEVA uses five cameras – one on each side of its rectangular body and one on the bottom – to spot cargo boxes and position itself precisely.

Once directly over a box, the robot:
1. Bends its legs to lower its body downward
2. Activates powered hooks that latch onto the top of the box
3. Secures the package beneath its body
4. Stands back up and continues on its journey

"It's similar to how forklifts work, but with much more flexibility," says mechanical engineer Thomas Schmidt, another member of the development team. "LEVA can go places no forklift could ever reach."

Finding Its Way Around

Like many modern robots, LEVA uses a combination of technologies to navigate:
- GPS for general location awareness
- LiDAR sensors that create detailed 3D maps of surroundings
- Cameras that help identify objects and read signs

This technology package allows the robot to travel through busy streets, avoid obstacles, and find efficient routes to its destination – all without human guidance.

From Laboratory to Streets

LEVA is currently a working prototype, with researchers still fine-tuning its ability to handle real-world challenges. The robot builds on technology from RIVR (formerly Swiss-Mile), an ETH Zurich spinoff company that already produces the RIVR ONE quadruped robot.

While RIVR ONE already operates as a delivery robot in some settings, LEVA's ability to handle cargo independently represents a significant step forward for autonomous delivery systems.

"Automating the entire delivery process has been a major challenge in robotics," notes Professor Martin Weber, who supervises the project. "Having robots that can pick up and drop off packages without human assistance could revolutionize last-mile delivery."

The researchers haven't announced when LEVA might be commercially available, but the prototype demonstrates that fully autonomous delivery robots capable of handling varied terrain and managing their own cargo are no longer science fiction – they're quickly becoming science fact.

As cities grow more congested and delivery demands increase, robots like LEVA might soon become familiar sights on streets around the world, quietly picking up packages and delivering them to their destinations without any human intervention needed along the way.