Drones join the fight against marine plastics

It’s one of Britain’s rare piping hot days; the sun is beating down and the Simpson-like clouds are floating majestically through a piercing blue sky. Widemouth Bay in North Cornwall is one of the many places across the UK where people, encouraged by the thermometer breaking into the mid teens, flood to on days like this.

Some people are donning their seemingly several-inch-thick wetsuits and paddling out into the freezing North Atlantic hoping to catch a wave, whilst others are content to stretch out on the fluffiest sand that the UK has to offer. What everyone has to endure however, is the enormous amount of plastic debris that can be found strewn across the beach.

70% of the planet is made up of water and one of the problems that research is telling us we need to focus more on, for our own well-being as well as the well-being of the planet, is the amount of plastic that is entering our oceans.

The University of Ghent in Belgium released a report earlier in the year suggesting that people who eat seafood on a regular basis can ingest up to 11,000 pieces of plastic in a year with some of these particles eventually ending up becoming embedded in our tissue.

Whilst the amount of plastic in the ocean is a recognised problem, little of it has been accounted for, with only 1% of the estimated eight million metric tonnes that enters our oceans every year, ever being tracked down.

The Plastic Tide initiative has been set up to help resolve this problem. Combining technology and painstaking crowd-sourced research, the project is using drones and machine learning algorithms to help shed light on an otherwise murky issue.


Earlier this month Peter Kohler, 33, and Ellie Mackay, 28, the founder and co-director of The Plastic Tide initiative, started the long trip that will see them travelling the length and breadth or Britain to gather data.

Kohler, who works on the project alongside world renowned oceanographer Dr Erik Van Sebille, explained the problem: “From Erik’s research we know that of all the plastics going into our ocean every year, we only know where 1% of it ends up and that is on the surface.”

“The other 99% is effectively unaccounted for and missing. So, it is either on the ocean bed, in the water column, in the marine animals or washing up on the beach where we are today in Widemouth.”

Having been issued my protective gloves, pick-up-stick and large, eco-friendly rubbish bag, I joined the dozen-or-so other volunteers and headed to the beach to commence trawling it for any foreign objects.

Just before we started the familiar harsh bleeps of a drone being powered up were heard, followed by, the aggressive buzzing that anyone who has heard one fly will recognise immediately. The drone proceeded to make slow paths up and down the beach, particularly along high-tide line, or tram line, which is where most of the debris can be found.


Ellie Mackay, the pilot of the almost £2000 DJI Phantom 4 drone, explained how they were using the latest tech to aid their research. “The technology is actually very straightforward. The ultra high-definition resolution of the camera on the drone is what makes what we’re doing possible and why this couldn’t be done five years ago.”

“What we do is fly up and down the beaches scanning the surface of the sand or pebbles taking thousands of images, these photos then get uploaded to an online platform called Zooniverse, which is a crowd-sourcing type system.”

“On this trip we are travelling 4,500 miles and expecting to take around 30,000 images. Once they are uploaded, the people who volunteer for Zooniverse will be able to tag the images thereby training the machine learning algorithm to determine what is plastic or rubbish and what is not, so that over time the computer will be able to tag the debris by itself.”

One of the main reasons for developing this research is to educate the public about the issue at hand, whilst also giving a greater insight into the worst effected areas of our coastline, and subsequently be able to more effectively plan how to deal with the issues. Mackay said, “every time we get that data it just adds more to the bigger picture”.

“We will be able to see temporal and seasonable changes and that might be to do with the weather and the tides or it might be to do with the humans that use the beach, and the more information we have, the more patterns we will be able to see within them. For instance, if the plastic is coming from the ocean we can then prepare properly if we know a storm is coming (more debris than normal is washed ashore during stormy weather) or we can track it back to where it originated. If it’s coming from beach users or other sources then we have to use an entirely different solution protocol. Currently we just do not have this capability and therefore we’re going about dealing with the problem very inefficiently.”

On their second day of surveying at the beginning of May, Kohler and Mackay were on a beach near Portsmouth and the amount of debris was astronomical. “We spoke to some of the local surfers who said during some of the summer months, the pollution was so bad that they simply didn’t want to go down to that spot despite it having the best waves. There is currently no explanation as to why that area is worse than other nearby spots and this is one of the answers we are seeking.”


As the sun was setting and the last surfers were shivering there way back up the beach, the debris-pickers volunteers gathered together at the drone’s temporary launch pad to assess our finds. A variety of debris had been collected, ranging from pieces of fishing nets, to bottle tops, to a Lego cutlass; which was presented to Jake, 3, who won the ‘ocean guardian’ of the day award. All the plastic was transferred into one bag and weighed. Although it only amounted to just over a kilo in weight, other beaches are not quite so lucky with Kohler saying on a clean they had recently been on, they gathered close to 16 kilos and considering plastic is quite a light material, this is an astronomical amount. (Click here for the report.)

One noticeable aspect of the pickup was how competitive people got and this is something else the Plastic Tide team want to harness.

“We’re teaming up with schools across the country,” Mackay explained. “The idea of collecting is very human. We are trying to utilise an app called the Marine Debris Tracker, which is an international platform, to encourage more young people to get involved. When you tag a piece of debris on the app it vibrates and gives you a similar sort of feeling to what you get when you get a Facebook or Instagram like. So if we can get young people engaged in this, like they were with Pokemon Go for instance and we can potentially team up with various companies to offer prizes, then it is a win win situation as they are having fun, data is being produced to develop the machine learning algorithm and the beaches are being cleaned.”

Having seen the project in action I wanted to speak to one of the men behind the algorithm and behind the idea of putting drones to use in combatting this ecological nightmare.

Dirk Gorrison is a specialist in utilising unmanned systems for humanitarian or conservation purposes. Whether tracking orangutans in the jungles of Borneo, fitting drones with ground penetrating radars to detect landmines or simply – or not so as the case may be – developing the algorithms to single out sea-plastic on Britain’s beaches, Gorreson is an expert in the field of computer science.

Whilst this kind of technology is not new, it is relatively speaking still in its infancy and many people are hesitant when delving into the realms of artificial intelligence (AI). Gorrison said: “Unfortunately when you say machine learning or AI a lot of people start thinking about evil killer robots that are going to take over the world and I die a little inside when I hear that.”

“You can’t blame them because that is the picture that the media portrays and people get stuck on a negative image. Also people are naturally defensive about things that they do not understand. The Plastic Tide algorithm is pretty simple though.”

“Essentially what we are doing is training the computer through repetition. If for instance the program was needed to tell the difference between a cat and a dog, if we tried to program the computer to recognise the difference through coding, that would be extremely complicated. However, if we enter hundreds or thousands of images into the computer, and by using a machine-learning algorithm, the computer will gradually be able to make a distinction between the two by itself. Obviously the more practice it gets the more accurate it will be.”

One thing the drones cannot do, for the time being at least, is clear up the rubbish and both Gorrison and Kohler were quick to highlight the importance of the volunteers who give up their time to clean the beaches. The day I spent in Widemouth demonstrated the willingness of local residents, and those on holiday, to look after and keep their beaches clean however Kohler said there is only so much one person can do. “We’ve been getting calls from people from the west coast of Scotland saying that the problem is so bad that they clean the beach one day and within the week just as much plastic has washed ashore again. Whilst it is great filling bags and seeming like you are doing your bit, it is quite disheartening when you are fighting a never-ending battle.”

Across the UK there are hundreds of organisations whose objective is to keep the British coastline clean, however one problem that Mackay outlined was that very few of them work together and as a result things are less efficient. “We hope that once we have an accurate map of the main problem areas, our platform will act as an umbrella for everyone else to pull together under, and plan their work accordingly.”

Although it remains one of Gorrison’s dreams to create an automated way of picking up the litter as well as tracking it, he is weary that this could have a detrimental effect. “I would like to make the process as easy as possible but the danger with getting robots or drones to tidy up the mess on beaches, is that people will become over reliant and ultimately not realise the scale of the problem. I have always been interested in how technology can interact with humanitarian and conservation projects, but I also believe that technology will ever beat humans. It’s merely one of the many tools that we have at our disposal in order to help solve the problems.”

The Plastic Tide team recognise that it is going to be some time before they have any conclusive results and although they are already starting to notice some trends, the accuracy of the project is reliant on time, given that the more data they collect, the more telling the output from the maps they produce will be. “In the meantime, the more awareness we can generate during our trip around Britain, the better. We know people have bigger priorities in their lives but this is a problem that will only escalate if the public, in general, don’t take more responsibility over their usage of plastic products. Plastics are an essential part of everyday life but people really do need to start to recognise the destructive effect that, if not dealt with properly, plastics have on our environment.”

So next time you’re on a beach, wherever you are in the world, remember the damaging effect that plastic is having on our oceans and get involved by picking up the litter you find. If you are in the uk then in just a few minutes you can tag what you have found and help develop the machine learning algorithm, by following this link.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: