is a startup I co-founded a year ago with 2 other UCSD engineers.

We are working with The Port of San Diego and The Scripps Institution of Oceanography to create the SeaSkimmer, an autonomous surface marine drone that cleans floating trash from ports and harbors.

here's our story:




in the port:

During the summer of 2018, we began to contact and interview relevant city and port officials to identify their pain points. We continued this process for several months and we learned in depth about The Port of San Diego's daily operations, its main costs, growth opportunities, and challenges.


We learned that The Port of San Diego spends a fortune to maintain water quality and avoid fines from increasingly stringent regulations. Water sample testing can take days, and officials often cannot react quickly enough to prevent contamination from spreading. In 2019, the most common method of floating trash cleanup is manually scooping it using nets, however this method is ineffective and expensive on a broad scale.

around the world:

Every minute, an entire garbage truck's worth of trash gets dumped into the ocean worldwide. This pollution ruins ecosystems, kills wildlife and eventually moves up the food chain,  poisons, and starves people. 

From our research, we realized that once trash has entered the middle of the ocean, cleanup is no longer logistically viable for reasons ranging from the low trash density per area, to the difficulty of transporting the trash back across the ocean, to the question of which countries are held accountable for this mid-ocean trash.




The only viable place to stop this trash crisis is at the source: In cities' ports and harbors, where the most garbage enters the water.

With our technical skillsets, my co-founders and I are well-equipped to develop SeaSkimmers that autonomously clean the floating garbage in ports and gather data which we can further analyze and eventually monetize. Our business simultaneously helps the port's bottom line and the pollution crisis.

Because our trash cleanup service is autonomous, it is logistically easy for the ports. Our solution is cheap and scaleable. and ports can now afford to fight the pollution crisis directly.



I adapted ArduPilot, an opensource software intended for autonomous flying drones for the SeaSkimmer's navigation.
The boat is currently capable of remote control and GPS waypoint navigation. We are currently working to implement our trained object detection model onto our boat using an onboard NVIDIA Jetson graphics processing unit.


For the last half year, we have been designing, manufacturing, and testing the first iteration of our drone. The boat has fiberglass pontoons and uses components from a treadmill as its conveyor belt to pick up floating trash.


I designed a few options for our concept models. I tried to strike a balance between imagination and reality when making these.

The majority of these models use a tri-hull design. The two small front pontoons provide ample space for trash intake, and the large back fuselage has a large storage volume.


The boats will have solar charging onboard, and depending on the use-case, may not need another power source. This self-sufficiency would drastically simplify the system, as each boat could have a self-contained, sealed electrical system.

The boats will have an onboard trash compactor/shredder. Since the trash is floating, by definition, it is not very dense and it can be compressed so we can store more. Solar-powered trash compacting garbage cans are a widespread and reliable technology which we can apply here.

In the future, we hope to make 'hubs' where the SeaSkimmers will dock, charge, deposit their garbage, and occasionally be serviced. The hub is modular so the system can easily scale to any port. The method of docking is important, depending to what degree of autonomy we will reach, and whether we expect the SeaSkimmers to charge while docked. We may use inductive. Their form factor is an important consideration, as the ports may want to place these hubs and SeaSkimmers in cramped spaces.