What litter really does to our neighbourhoods
A can by the roadside, cigarette butts at a bus stop, or fast food packaging in the park: litter may seem trivial, but it could affect far more than we think. Through a new citizen science project, UAntwerp and national daily De Morgen not only want to map litter hotspots, but also investigate what a dirty street does to us and what solutions there are. Citizens will be able to register litter through a smart app and help train the AI model that analyses the images.
In short
- In 2027, UAntwerp and De Morgen will launch a large-scale citizen science project on litter.
- Through an app, citizens will be able to upload photos of litter while also answering questions about how they experience their neighbourhood and what actions they expect.
- The interdisciplinary research combines social sciences, health sciences, environmental policy and artificial intelligence.
- Not only do the researchers want to measure litter, but also to understand what litter does to wellbeing and social connectedness.
Everyone has an opinion about litter
‘Litter is one of the biggest complaints amongst citizens,’ says Jasna Loos, research manager at the Faculty of Medicine and Health Sciences. ‘Whenever you bring up the topic, almost everyone immediately starts sharing a personal story.’
Kris Bachus, associate professor of environmental policy (Faculty of Social Sciences), recognises that too: ‘There’s a lot of frustration about litter in Flanders. At the same time, we still know too little about what it concretely does to the way people experience their environment and what they expect from policymakers.’
This forms the core of the new research project by UAntwerp and De Morgen. ‘The project will not only look at the impact of litter on wellbeing and community spirit, but also at the complex factors that lead to litter problems in certain neighbourhoods, and how these can be addressed locally,’ Loos adds.
Citizen Science at the Nerdland Festival
The idea grew out of a request from De Morgen to the university. ‘The newspaper wanted to set up a citizen science project together,’ Bachus explains. ‘A university has methodological expertise and scientific knowledge, but it doesn’t easily reach tens of thousands of citizens. That’s something a media partner such as De Morgen can do. That makes us highly complementary.’ Loos also sees major added value in this: ‘Citizen science makes research more relevant to society. When people help collect data themselves, they’re also more likely to actively engage with the results afterwards.’
‘This time, we want to involve citizens earlier in the process than is usually the case,’ says Loos. Even before the official launch, the research project will have its first soft launch during the Nerdland Festival from 22 to 25 May. The researchers will take part in a Focus Talk on urban mining, the newspaper and the university will host a stand with information about the project, and visitors will be able to upload photos of litter and help train the AI model. ‘We’ll also literally ask: which questions should we definitely not forget? What issues matter in your neighbourhood? Are our research questions formulated correctly? That input is hugely important,’ says Loos.
AI model for cigarette butts, plastic and cans
Siegfried Mercelis, lecturer at the Faculty of Applied Engineering and affiliated with imec-IDLab, is responsible for the technological side of the project: ‘Initially, we’ll mainly collect photos. That allows us to build a dataset we can use to train the AI model. The goal is for the technology to automatically learn to recognise different types of litter.’
There’s a lot of frustration about litter in Flanders. At the same time, we still know too little about what it concretely does to the way people experience their environment and what they expect from policymakers.
That sounds easier than it is. ‘Recognising a can is usually fairly straightforward,’ says Mercelis. ‘But wrappers or partially covered packaging are much more difficult. Take a cigarette butt between autumn leaves: the model has to learn that only the cigarette butt is relevant, not the leaves around it.’
Lighting, rain and the angle of a photo also play a role. ‘That’s why at Nerdland, we want to test how people take photos: from close up or far away, under which conditions and in what lighting. That helps us make the system more robust.’ The final app should be ready by spring 2027. Users will be able to upload photos and add labels to the visible litter. ‘We’re basically building a kind of ObsIdentify for litter,’ Mercelis laughs.
More than a problem of cleanliness
For the researchers, the project is about far more than just clean streets. ‘Litter touches on rights and responsibilities,’ Bachus believes. ‘People feel frustrated because they think others aren’t taking responsibility. That irritation is often less about the environment and more about liveability: people want their surroundings to remain pleasant and attractive.’ Loos also emphasises the social impact: ‘Public space affects our mental wellbeing. If an environment feels dirty or neglected, that also affects social connectedness and, by extension, our overall health.’
The researchers also want to investigate whether some groups are more sensitive to that impact than others. ‘We want to understand that complexity,’ says Loos. ‘What kind of people live in a neighbourhood? Do people feel jointly responsible for cleanliness, or do they see it as a task for the government? That can differ culturally and socially.’ Researchers from the Stadskracht (Power of the City) expertise centre at AP University of Applied Sciences and Arts will therefore also set up an in-depth case study involving interviews in different neighbourhoods.
From cigarette butts to circular economy
The research also aims to look beyond simply cleaning up. ‘We want to broaden the debate towards the circular economy and reuse,’ says Bachus. ‘If you succeed in reducing the use of disposable packaging, you automatically reduce the risk of litter as well.’
The researchers already know the most common forms of litter. ‘Cigarette butts are number one by far,’ says Loos. ‘They’re followed by drink cans and single-use plastics, such as plastic bottles or straws. Cigarette butts contain thousands of toxic substances, and most litter breaks down into microplastics that end up in our soil, water and animals. Through the food chain, they even end up in the human body. The One Health perspective — the idea that the health of people, animals and our wider ecosystem is interconnected — also applies here.’
Can this behaviour change?
The researchers hope the project won’t only generate data, but also raise awareness. ‘I hope that it will help people to better understand the wider effects of litter,’ says Loos. ‘And that, as a result, they may also take action themselves more quickly and help keep their neighbourhood clean.’
According to Bachus, that’s precisely the strength of citizen science: ‘The people themselves produce the data. That makes the project both unique and vulnerable. But experience shows that participants are hugely motivated. They want to know what’s happening in their neighbourhood and feel involved. Ultimately, it all comes down to the bigger question: how do we want to live together in our public space?’
Want to stay updated on this and other research?
Subscribe for UAntwerp's newsletter, full of science news, lectures and events.