In two cities in eastern India, Pamela Dalton's team walks around pointing Nasal Rangers - devices resembling oversized hair dryers - into half-completed community toilets.
Then they sniff.
Dalton is an experimental psychologist from Philadelphia whose specialty is how people perceive and respond to odors. The odd-looking devices collect chemical data on aromas of all kinds, before and after the toilets are open for use. The goal: Get more people to use the facilities.
People don't want to relieve themselves indoors, Dalton said, and the intensity of bad smells is part of the problem. While the smell of human waste is diluted outdoors, without proper sanitation, it concentrates indoors, sending residents to relieve themselves elsewhere.
Poor sanitation is a leading cause of child mortality and disease in the developing world. India has the highest rates of what officials call open defecation, according to the World Health Organization. Many residents of urban slums come from villages where they may never have seen a modern toilet, and have no idea that waste can be infectious.
"People just defecate wherever," said Dalton, a faculty member at the Monell Chemical Senses Center. "They're used to going outside."
So Dalton's team, which is overseeing construction of dozens of new community toilets, is trying to make the new facilities more appealing. Part of that is studying the current state of stench.
"What we're doing with odor evaluation is one piece of behavior change," she said recently after returning home from India.
The work is part of the Potty Project, a vast effort by researchers, architects, nonprofit organizations, and contractors, commissioned by the Bill and Melinda Gates Foundation five years ago to eliminate the human waste problem in five Indian cities.
Western attempts to improve living conditions in developing countries often focus on infrastructure and technology, and they often fail. The Potty Project includes a significant amount of research to understand what motivates people to use the bathroom indoors and to keep themselves (and the facilities) clean.
Dalton has conducted related research for years. In the late 1990s, in response to complaints about the smell from a wastewater treatment plant in south Camden, she used olfactometers to track the aroma scientifically, interviewed worried residents about their health and well-being, and tested the residents' sensitivity to smell. Improvements were made to the plant.
She said the Gates foundation asked for her help understanding the odors in Indian slums and how residents react to them.
The Indian government has a target to make the nation "Open Defecation Free" by 2019. Organizations such as UNICEF are stepping in to help with a problem that goes far beyond stink.
Going outdoors "has terrible consequences for health," said Cristina Bicchieri, a professor of social thought and comparative ethics at the University of Pennsylvania who studies judgment and decision-making.
It's also a safety issue, she said. "It's women who want latrines. It's unsafe for them to defecate outside."
Solving the problem is complex.
The teeming slums don't have the plumbing and septic systems needed to accommodate all the people who arrive from rural and tribal areas looking for work.
Urban slums are "more mobile and transient" than other types of communities, said Radhika Nagesh, a research associate working with Dalton and Project Sammaan, the local group involved with the sanitation effort. "It's difficult for permanent or household-level sanitation services to be set up."
Currently, residents may have to travel a mile to reach an indoor toilet.
"It's not that the structure doesn't exist," Nagesh said by Skype from India. "It's that it never gets cleaned. Parts get broken and never repaired. And there's no management structure in place that will make these a long-term solution."
The team is building 58 community toilet facilities - each has multiple commodes and some have showers and clothes-washing stations - in the slums of Bhubaneswar and Cuttack, neighboring cities in the eastern state of Odisha. There will be a small fee to use them, with proceeds going to keep the septic systems clean.
Unlike the current facilities, these will be managed and maintained by local residents, a plan that is intended to give them a sense of ownership.
Dalton's crew of a half-dozen surveyors armed with sniffing devices (plus data entry and other support from local workers) has been monitoring the air during construction. They will continue after the toilets open and, they hope, become heavily used.
Data collected by the commercially available Nasal Rangers, technically known as field olfactometers, can be entered into an app to track the intensity and location of odors at different facilities and times of day to spot patterns. The team will also use special fibers to collect microscopic substances that carry odors. Those will be sent back to the lab at Monell.
"If any of the cleaning products used actually mask the odor," Dalton said in an email, the bad-smelling compound "may still be present but not perceived."
Until, perhaps, a change in cleaning practices inadvertently exposes the stink.
The first 20 or so blocks of toilets are scheduled to open shortly. Meanwhile, the trained team has been visiting the sites to record the odors in the area that are already there: foods cooking, waste rotting, sewage sitting. When the new toilets open and emit odors, the team will be able to take into account what the place already smelled like.
Dalton, who in recent months has been monitoring progress from afar, just booked flights to India for December so she can get a whiff herself.