This project is a work-in-progress!


Plastic pollution in rivers and ocean is a big issue:

  • Between 1.15 and 2.41 tonnes of plastic waste currently enters the ocean every year from rivers. 
  • Plastic pollution is long-term as they are durable in the environment 
  • Plastic pollution poses threats on aquatic life, ecosystems, and human health. 

A key step in stopping ocean plastic pollution is to tackle the problem beginning in rivers.

Current solutions for river plastic pollution require large manpower and/or cost and low efficiency:

  • Large machines with nets intercepting the pollution (for example the Ocean Cleanup Interceptor):
    • Effective but extremely costly; not worth employing in smaller rivers
  • Manual cleanup using boats and nets:
    • Not very effective; requires large manpower
  • Small manual remote controlled boats:
    • Not very effective and nonspecific to trash; mainly in harbors and small lakes

What is this?

The goal is to create a boat that turns and captures trash on its own in even small waterways, thus autonomously cleaning up rivers.

HydroCleaner path versus conventional remote controlled path. The HydroCleaner path “simulates” near-human control while being completely autonomous and cleans an area of waterway much faster than a conventional robot, which needs to sweep all areas without image guidance.

Mechanical Design

  • Catamaran hull: two pontoons, two motors, propellers, ESCs.
  • Joined by a 660mm aluminum extrusion with electronics mounted. 
  • Net on back of boat to collect trash.
  • As the boat moves forward, trash flows in between the two pontoons into the net. It should not reverse.

Machine Learning

The boat is powered using a Jetson Nano, details found in the log. But the main idea is that the boat would navigate on its own via GNSS guidance, and whenever it spots a trash using the camera, it will automatically navigate to the object.