Every year, millions of tonnes of trash made of plastic float down polluted urban rivers and industrial waterways and into the oceans. Now, a company based in Hong Kong has come up with a way to stop these dangerous flows of trash.
Open Ocean Engineering has made Clearbot Neo, a sleek robotic boat with artificial intelligence that collects tonnes of floating trash that would otherwise wash into the Pacific from the busy harbour in the territory.
After a long time developing it, the people who made it want to make it bigger and have fleets of Clearbot Neos clean and protect water all over the world.
And there are worries that by 2040, the amount of plastic trash going into the oceans could almost triple, adding between 23 and 37 million metric tonnes per year. That would be about 50 kg of trash made of plastic per metre of coastline around the world.
“If we clean up our rivers and harbours, we help clean up our oceans,” says Sidhant Gupta, one of the people who made Clearbot Neo.
The Clearbot Neo is only three metres long and is moved by an electric motor powered by a solar battery. It moves up and down certain sections of water in a systematic way, similar to how a robot cleaner moves across a living room floor.
The Clearbot Neo is perfect for use in harbours, canals, and rivers because it is small and doesn’t take up as much space as other solutions for picking up trash on the high seas.
It floats on the water’s surface and picks up floating trash with a conveyor belt built into its bow between its two hulls and a holding bin built into its back.
Clearbot Neo uses AI to figure out what kinds of trash it picks up and where it came from.
It can take in up to a metric tonne of trash per day to be recycled or thrown away. With a custom-made boom, it can clean up oil and fuel spills in a small area by collecting up to 15 litres of pollution a day.
But this isn’t just a simple machine for cleaning. Using a two-camera detection system, it also stores a lot of information in the cloud.
One camera scans the surface of the water so that the bot can find trash and stay away from marine life, navigational hazards, and other boats. This makes it safe and useful for work in rivers and harbours.
The second camera takes a picture of every piece of trash that lands on the conveyor belt and sends its image and GPS location to the company’s data compliance system, which is hosted on Microsoft’s Azure platform.
Here some Related photo
When this information is combined with other variables, like information about sea currents and tides, environmentalists and marine authorities can start to figure out where the trash is coming from. Data about the quality of the water is also sent to the cloud.
After graduating from Hong Kong University in 2019, computer engineers Gupta and Utkarsh Goel started their own company and started working on their Clearbot solution right away.
When they went on vacation to the Indonesian island of Bali, they saw how local workers would take to the water every day in small boats and even on surfboards to manually fish trash out of the sea to keep the shoreline and beaches safe and clean for tourists.
That made the two partners wonder: How could this slow and difficult process be made easier?
Gupta and Goel made a simple aluminium prototype in Bali, and when they got back to Hong Kong, they improved it by making it out of fibreglass. After that, a number of prototypes were made, the most recent of which is the sleek Clearbot Neo.
The hardest part of the project was making an AI model that could find waste in the water and figure out what it was.
Gupta says, “We just didn’t have enough computer power to train, run, and test the models.” “Exactly. This is where Azure comes in. In the end, Microsoft gave us a grant for AI for Earth in the spring of 2020, and we spent the next year building the AI model on Azure.
“It took a while because we didn’t have enough data to train it well at first, but we were able to build a model very quickly. Then we put it on the robot and started teaching it to plan routes, pick up trash, and make data.”
With the help of GPS, Clearbot Neo can both clean up the trash and make a data point for each item it picks up. This data point includes the location, size, type, material, and weight of each item it picks up. After every mission, Azure’s AI has already put the Clearbot Neo’s haul into categories and added it to a database that is growing.
We are trying to figure out how trash gets into the water in the first place.
Gupta says, “We use Azure Functions, Azure Container Registry, and Container Instances to help us run our back-end systems and deploy our AI models.”
In fact, the hard data is more important than the physical things that are being collected. Only 20% to 40% of the plastic trash in the water around Hong Kong can be recycled. Most of it is dirty or broken down too much to be useful.
But Gupta says that Clearbot Neo and Azure are helping them figure out how trash gets into the water in the first place. “It makes the process of cleaning up the oceans much more clear. We collect information about what’s in the water, what it’s made of, how much of it can be recycled, and what materials we should focus on.
Clearbot Neo and its customers (companies, governments, non-government organisations, and individuals) don’t have to guess how big the problem is in their waters because they have a lot of information. They can find out where marine waste comes from and then stop pollution where it starts.
A Hong Kong real estate company called Sino Group recently bought a model. They will work with the Clearbot team to help keep a yacht marina clean and test the technology even more.
“Sino Group is committed to promoting sustainable living and green operations in the communities where it works, and it wants to find green and sustainable solutions for the real estate industry,” says Andrew Young, associate director of the company’s innovation department.
“We think the electric-powered Clearbot is a good way for the marina at Gold Coast Yacht and Country Club to automatically collect trash from the surface of the water. Its artificial intelligence (AI) and ability to find its own way are also good features. It is a green solution that doesn’t make noise or hurt the air. So, we’re happy to work with Clearbot by letting them use the marina as a place to test the solution.
So far, the Clearbot Neo has only been used in the water around Hong Kong. Gupta and Goel have been contacted by several potential customers in other countries. However, travel restrictions due to the pandemic have made it hard for them to grow their business right now.
Gupta and Goel are sure that they have a global solution, and they see a future where fleets of Clearbot Neos work together to patrol waterways all over the world.
“The whole problem,” says Gupta, “is that there aren’t enough resources to deal with pollution.” Why is this the case? Because it doesn’t make any money. Right now, you can add that, but that’s another story. So that’s where we stand. I’m really happy that we can do this work and make money from it. We can take a broad idea and turn it into something very specific that works.”
