Urban "smellscapes" of London and Barcelona have been created in the hope that city planners will take more notice of inhabitants' olfactory experiences when it comes to future decisions and designs.
"Humans are able to potentially discriminate more than one trillion different odours, yet city officials and urban planners deal only with the management of less than ten bad odours," Daniele Quercia, co-author on a paper describing the tool, told WIRED. "Why this negative and oversimplified perspective? Smell is simply hard to measure. Cities are victims of a discipline's negative perspective. The goal of our work is to open up a new stream of research that celebrates the positive role that smell has to play in city life."
The resulting urban scent dictionary and maps are the work of Cambridge University's Quercia, two other computer scientists – Rossano Schifanella of the University of Torino and Luca Maria Aiello from Yahoo Labs – and Royal College of Art designer Kate McLean. Quercia, Schifanella and Aiello had already collaborated on a walkit.com-style navigation tool that takes people along the most "emotionally pleasant" route. "Participants loved the pleasant directions but also expressed the need to integrate smell into the recommendation process," Aiello says. "That is why we teamed up with Kate, who is an olfactory researcher."
McLean had been operating "smell walks" for years across Glasgow, Edinburgh, Amsterdam, Pamplona, Paris, Newport and New York, where she asked locals to wander around their city and note down sensory impressions. "I deconstructed the city experience into five sensory modalities and mapped each one separately," McLean continued. "Smell was the most captivating, and also significantly under-researched, so I decided to continue. I was attracted by its difficulty to pin down, its ephemerality, its multiple temporalities and also its subjectivity and capacity to generate time travel and momentary location-displacement on the part of the sniffer. In short, it was problematic and that is what makes for a great art/design research study."
The words she captured were collected by the wider team and cross-referenced with geotagged social media posts and photos on Flickr (530,000), Twitter (113,000) and Instagram (35,000) from London and Barcelona. Words were split between pleasant smells – including "baked goods", "cut grass", "afterhave", "woodlands", "petrol", "soap powder", "babies" and plenty of flower names – and unpleasant smells, including "flatulence", "faeces", "smoke", "urinals" and "fumes".
These were then grouped into categories by looking at how often different words were used in the same post – so if cigarette and tobacco were used in the same sentence repeatedly, they could be considered as originating from one category. "With this assumption, we ran an algorithm to automatically detect clusters of words often occurring together, and we found that the emerging clusters [had a] striking resemble [to] the smell categories from the literature," Aiello says. The literature he refers to is described in detail in the team's paper, entitled "Smelly Maps: The Digital Life of Urban Smellscapes", and includes past work by the late urban planner and academic Victoria Henshaw.
Scent-word analysis in hand, the team then looked to the perfume industry to refine their olfactory rhetoric – top notes being the instantly recognisable smells that hit a person as soon as they walk down a street but dissipate just as fast; base notes taking a while to get going but likely to linger for hours. "City odours work exactly the same!" Schifanella says. "The base notes for Barcelona and London are emissions and nature." Those scents blend with locally dominant scents – like the animals near London Zoo – to produce "middle notes". "The result of this process is the first urban smell dictionary containing 285 English terms," Quercia says.
The team also created "smell clusters" and compared these with city air quality indicators. They found there was plenty of correlation between the two – in geographical regions where plenty of emission-related words were used and few nature-related ones, the air quality had been categorised poorly by city officials.
Quercia hopes that beyond this, the work will encourage researchers to study positive as well as the more commonly investigated negative city smells we all know so well. "If they were to do that, there would be a number of ways that the urban smellscape could be altered. A few examples include changing the street layout [to modify the airflow], pedestrianisation [to lower traffic emissions], and planting trees [to create restorative environments]."
He also imagines new tools that offer olfactory routes through the city for runners – anyone can build on the basic principles of their work to create techniques for interpreting city smells. "If we were to design an app for collecting odours, that app could only be used to collect odours in specific parts of town at specific moments in time," Quercia explains. "That way we would have high spatial and temporal resolution but low geographic coverage. The app would be able to collect the top notes that social media is not able to capture." "I hope that urban planners and architects could use my mapping methodologies as a way of recording and collating their own data," McLean says. "But what I have discovered inadvertently in the course of my research is how smell does make a difference to how people think about where they live."
"Smell gets a bad rap, often perceived as a negative, but there are many surprisingly good smells out there if we pause to notice them. In the Western world we are trying to eliminate smell but until we actually have the discussion as to the smells that we appreciate and recognise as belonging to a place, why should we accept the elimination of odours?"
This story originally featured on WIRED in June 2015.
This article was originally published by WIRED UK
