Ideas | Justin Chen

Justin Chen: How robots turn sewers into sensors

Sewer pipes stacked on Memorial Drive near Amesbury Street in Cambridge.
Sewer pipes stacked on Memorial Drive near Amesbury Street in Cambridge.Pat Greenhouse/Globe Staff/Globe Staff

The French novelist Victor Hugo described the sewage system as “the consciousness of the city,” a place where there are no secrets. Transporting millions of gallons of wastewater, the sewer contains a record of the public’s health and the substances it consumes.

Today, science has made possible what Hugo could not have fathomed in his day: water-sampling robots placed at strategic points in a sewer system and capable of delivering ever-more precise information about a community’s health.

As the country confronts an opioid crisis that kills more than 60,000 American each year, one Cambridge-based company is hoping that it can use that kind of technology to measure traces of the drugs in sewers. Doing so, according to the firm, Biobot Analytics, could help to reveal remarkably detailed patterns of drug use — and give communities a powerful tool to detect emerging public health threats.


“Our goal,” said Newsha Ghaeli, one of the company’s co-founders, “is to transform sewers into public health observatories.”

The question is whether communities are ready to reveal their own secrets. Researchers, using wastewater-based epidemiology, would effectively be treating sewage as a collective urine and stool sample, made available willingly — or not — by everyone who uses a toilet. And by looking for the specific molecules that are found when opioids pass through the human body, it would be possible to distinguish between drugs that are flushed down a toilet and those that were ingested.

“It really comes down to finding a city that wants to know about what’s in their sewer,” said Daniel Burgard, chair of the chemistry department at the University of Puget Sound.

With the stigma attached to drug addiction, some cities may be reluctant to know more about the extent of local opioid use, Burgard said. Government officials and residents may see wastewater-based epidemiology less as a public health tool and more as a threat to their own reputations.

Biobot Analytics will also need to find customers willing to pay hundreds of thousands of dollars to use its technology. So far, Ghaeli and her co-founder, Mariana Matus, both former MIT researchers, have tested their approach to wastewater-based epidemiology in Boston, Kuwait, and Seoul, using their robots to detect bacteria.


Many European cities have already embraced wastewater-based epidemiology techniques to detect illegal substances in their sewer systems. But such testing is conducted at wastewater treatment facilities.

Analyzing wastewater before it mixes downstream at treatment facilities permits Biobot to measure drug use of not only an entire city but also of specific locations, down to areas of a few thousand people. The more precise method could enable city officials to first pinpoint communities that need interventions, like substance abuse programs, and to later measure the success of those programs in lowering drug use.

“Going down into the sewer system below the city is the most valuable approach compared to just looking into the influx of a big wastewater treatment plant,” said Carsten Prasse, assistant professor in the department of environmental health and engineering at Johns Hopkins University.

Biobot’s technology will get a test run in Cary, N.C. Located near North Carolina’s Research Triangle, the town of 160,000 may seem like an unlikely proving ground. It has one of the highest median household incomes in North Carolina and one of the lowest crime rates in the country. It recently hosted a festival where participants sculpted mice out of pimento cheese; town officials plan to celebrate International Yoga Day with free classes.

But like other communities, Cary has learned that it’s not immune from the opioid epidemic. While searching for ways to deal with the crisis, municipal officials learned about wastewater epidemiology and contacted Biobot.


Together, they designed a three-month pilot program that is expected to run through this summer. The project will be funded by a $100,000 grant from Bloomberg Philanthropies as part of a nationwide competition that encourages town leaders to address problems with innovative ideas.

Biobot and Cary will install robots in 10 locations. Over a 24-hour period, each robot will pump 2½ gallons of water through a filter, trapping any dissolved chemicals.

Biobot scientists, with the help of a machine that identifies molecules based on their mass, will then analyze the filter to estimate consumption of legal and illegal opioids including morphine, hydromophone, oxycodone, heroin, and fentanyl. The sampling and subsequent analysis will be performed once every two weeks to follow drug use over time.

Because each of the sample areas includes data from 4,000 to 16,000 residents, scientists will not be able to track individual drug use.

Biobot is hoping that success in Cary will help convince other cities to adopt its technology.

Europe has seen success in testing based on wastewater treatment facilities. The European Monitoring Center for Drugs and Drug Addiction, an agency of the European Union, collates the data collected at those facilities. In 2018, it described drug use in 56 cities located across 19 countries.

Experts believe that multiple factors contribute to Europe’s early acceptance of wastewater-based epidemiology. European governments play a more active role in health issues, making city officials and citizens receptive to the practice. And Europe has developed a critical mass of scientists and government officials sharing data. Wastewater-based epidemiologists in the United States are hopeful that a similar critical mass will form in the next few years.


Justin Chen is a 2018 AAAS Mass Media Science and Engineering Fellow and a reporting intern at STAT, from which this piece was adapted.