Indian low-cost AQI startups AMBEE, AirVeda’s certification up in the air

Sarita Ahlawat holds up her startup Living Sciences’ air quality monitor against a cold, grey December sky. Its PM 10 reading—particulate matter measured in micrograms—flickers before settling on 263. As we carry it out of her office, to the foyer, and then to the balcony, the reading fluctuates, before it finally settles at 313. According to the National Ambient Air Quality Standards (NAAQS), we’re breathing in “very poor” air. 

“Ideally, we shouldn’t even be sitting in this room,” she says. All around her are students who help her manufacture and collect data from low-cost air quality monitors.

Ahlawat and her team have installed 30 monitors across Delhi’s Indian Institute of Technology, but want to hit 100 in the coming months. “We don’t realise how pollution levels change even with a 300-acre campus,” she says.

The Living Sciences air quality monitors sell for $10-$15, their demand increasing faster than Ahlawat and her team can make them. “Corporates even give them away as gifts,” she says. By the end of 2019, Living Sciences will rake in Rs 50 lakh ($70,262) from the sales. Not bad for a two-year-old startup. 

The demand is driven by the lack of hyperlocal data that can accurately measure where, and why, air quality has dipped. And this in a country that is home to 14 of the 15 most polluted cities in the world. “Hyperlocal data can help identify and deal with hotspots such as parking lots, where loose dust needs to be controlled. Even simple things like knowing when not to work out in the open is beneficial,” says Ahlawat. 

At a national level, the recommended density for air pollution monitors, as per the Central Pollution Control Board (CPCB), is 4 per 100,000 people. CPCB also claims India needs 4,000 official monitors to get an accurate representation of air quality. 

The current count? 793.

A new wave of engineers, academics and electronic equipment makers such as AMBEE, Kaiterra, and AirVeda are now trying to plug these gaps. They provide low-cost, personalised air quality monitors, ranging from as little as Rs 2,000 ($28) right up to Rs 10,000 ($141) on Amazon. Schools, offices, residential highrises all use a network of these monitors to track indoor air quality. The monitors even come with a personalised air quality dashboard, connected to an app. In addition to selling hardware, startups like AMBEE are now also engaged in analysing and predicting the Air Quality Index (AQI) of unmonitored places.

Growing at a compounded annual growth rate (CAGR) of 4.6% globally, the nascent air monitoring market will be worth approximately $5.5 billion by 2024, according to a market research report. The market includes all types of monitoring equipment, but continuous and indoor air monitoring systems have the largest share. 

Personal monitors can help dial back the pollution levels indoors by telling you, for example, whether to keep that window open or not. Outdoors, they can help map areas that haven’t been monitored for air quality. At stake is the opportunity to mitigate India’s health burden caused by air quality that’s several notches below WHO recommended levels. According to a 2019 Lancet study, bad air quality has already shaved off 1.7 years from the average Indian lifespan. 

But there’s just one, rather big thorn in their flesh—the lack of any official standards to certify these monitors, and the data they generate. “These low-cost monitors have no approval from any certification body. There’s no way to know if their readings are accurate. Globally, no country has certified sensor-based monitors,” says Gurnam Singh, head of the Delhi-based Air Quality Lab at CPCB. Wear and tear, say experts, can throw readings off in low-cost monitors, and have to be checked regularly. 

CPCB’s skepticism, however, hasn’t exactly slowed down the startups. 

Hyperlocal data, hyperlocal action

Namita Gupta, founder of AirVeda, made her first DIY monitor at home in 2016, to track the air quality around her asthmatic daughter. Now, some of her main clients are schools. 

“Schools with monitors have centralised dashboards. They can share information and alerts with parents,” says Gupta. AirVeda’s devices sell for Rs 9,500 ($133) apiece, down from Rs 11,000 ($154) when they first launched in 2016. While Gupta doesn’t strictly prescribe the density of monitors, she recommends that schools follow some broad rules. “For instance, there should be one monitor in every type of classroom, or per 5,000 square feet,” says Gupta. 

The company also sells annual data plans for Rs. 7,000 ($98.4) per monitor, which could set a school back by Rs 3-4 lakh ($4,217-5,623). What may seem like a high upfront expense, claims Gupta, is actually a tenth of what schools spend on air purifiers.“There aren’t enough volumes yet to bring the price down further,” admits Gupta. Most monitors encase sensors and microprocessors imported from China, adding to the cost. Kaiterra’s building grade monitor—SensorEdge—retails for Rs. 80,000 ($1,124). 

AMBEE, an air quality monitoring and data sciences company, retails their mid-range API (Application Programme Interface) plan at $280. “Health insurance companies and fitness apps are now paying AMBEE to use their APIs to target potential customers. I never realised this could be a revenue stream for us,” says Maddy Anand, the company’s co-founder. 

Fine tuned

AMBEE has tapped companies keen to sell products that mitigate the effects of air pollution. One of his clients is a digital health company with 2.5 million daily active users, who get real time updates on where and when to work out in the open

These companies also sell plans to corporate offices. For instance, Noida-based Info Edge India, an online classifieds company, procured 15-20 devices—one for each floor of their three offices. “Once we were able to monitor indoor air quality, we integrated an electronic air cleaner in our central air conditioning system,” says Desh Deep Singh, a senior facilities manager at Info Edge.

These commercial and personal air quality monitors are sleek, portable (can be carried from room to room) and connect to monitoring apps. They work with an optical laser technology that shows ambient readings in real time. 

Government monitors, in comparison, are elaborate systems that measure parameters like PM10, sulphur oxide and nitrous oxide. But they are also stationary, rectangular sheds with clunky equipment. Often, manual readings have to be taken, which are released only once in 24 hours.

“The personal monitor can tell me if I need more purifiers, or if I can actually open up the window for some parts of the day,” says Namrata Garg, a Gurugram-based resident. Garg bought Kaiterra’s Laser Egg model that displays ambient levels of fine particles (PM 2.5), PM 10 and even carbon dioxide (CO2). 

Healthy indoor levels of CO2 should be between 300-1000 parts per million (ppm). “2000 ppm is when you start to get drowsy, like when you’re in stuffy classrooms or conference halls,” says Nita Soans, the India representative for Kaiterra. The company’s models that depict ambient CO2 levels are pricier by Rs 3,000 ($42) than the basic ones, which only show PM 2.5 and PM 10 readings.

“When fitness goals, food and shopping, everything can be personalised through an app, why not the ambient air quality around you?” asks Anand. AMBEE claims it’s going a step further by using various metrics —readings from their own monitors, wind speed, traffic, precipitation, satellite data etc.—to create a real-time AQI map for “every zip code in the country,” claims Anand.

Hybrid hope 

The magic with these commercial monitors lies in the algorithms, explains Ahlawat. The monitors’ optical sensors capture the particles passing through them, and record their size. “The algorithm then uses the size and pattern the dust particle makes on the chip to map it to the air quality index (AQI),” she says. 

Imported sensors have to calibrated—or adjusted—to Indian conditions. For instance, says Ahlawat, the size of dust particles in India is different from the US. 

With proper calibration, says Gupta, a network of low-cost monitors can geographically cover areas that either fall between two government monitors, or aren’t covered at all. A hybrid system like this can both substantiate data and help build a more accurate air quality profile of a city or a town, she adds. 

While Delhi’s official monitoring system now has approximately 55 stations that either manually or automatically record air quality data, other metros like Bengaluru and Mumbai have fewer—19 and 5, respectively. 

AMBEE, Kaiterra, AirVeda, and Living Sciences have all deployed their own small networks of monitors to source localised information on outdoor air quality. AirVeda is doing this in partnership with Gurugram Municipal Development Authority (GMDA), which has invited startups to install monitors at major intersections. 

For AMBEE, their core business pivoted in 2016. From just selling monitors, they now build prediction models for unmapped smaller towns and cities like Tumkur, Chitradurga, Manali or Kodaikanal. It is also a service provider on Smart City projects—the central government’s ambitious, urban renewal plan for Indian cities—for electronics major BOSCH. Through this, Anand is hopeful AMBEE will get more data to refine the in-house algorithms and improve AQI predictability.

Between AMBEE and AirVeda, a few smaller state government projects have been triggered. But there is a particularly tough client-collaborator standing in their way—the CPCB.

Keep CAAQM and move on 

The CPCB has a steep mandate to cut pollution by 35% in India’s non-attainment cities in the next three to five years. Non-attainment cities are the ones that violate the NAAQS based on air quality data collected between 2011-15. And cities that don’t have an action plan will be pulled up and charged penalties up to Rs. 5 lakh ($7,029) per month.

So technically, it makes sense to augment the patchy air monitoring network with low-cost monitors. 

Except, the Board has no use for “unofficial” sources of air quality data. Despite GMDA’s current tie-up with AirVeda, CPCB’s Singh is clear that until such certification comes through, a widespread hybrid network of official and unofficial monitors can’t co-exist.

“The best they can do is give us an indication of air quality in an area. They won’t give exact figures,” he says.

The biggest challenge with low-cost monitors, admits Kaiterra’s Soans, is that they “drift”. Exposure to too much dust, moisture, extreme weather conditions, or just the passage of time, could throw their readings off.  They also need to be checked against readings from official monitors to make sure they’re accurate—there is usually a deviation of 10-15% in the data.

The startups do this process, called co-locating, by comparing their readings against Beta-Attenuation Machines (BAM) that have been approved by the United States Environmental Protection Agency (US EPA)—a certification that CPCB accepts. 

BAMs look like old-fashioned cassette players. The filter tape captures particulate matter. BAMs are used to measure PM 10 and PM 2.5 levels (Picture credit: Olina Banerji/The Ken)

Increasingly, the CPCB has moved away from providing just 24-hour averages to continuous monitoring and data dissemination online. Their multi-equipment stations use air samplers, which capture particulate matter and weigh them. The filters in these samplers are then analysed physically to determine levels of sulphur oxide, nitrous oxide, CO2, nickel and lead, among other gases. Regulatory-grade samplers come certified by the US EPA and are considered to be the most accurate benchmark to measure against. 

There’s another thing that this piece of bulky government machinery can do that commercial monitors can’t—attribute sources to the pollution. “Hotspot data has limited value in actually mitigating air pollution, unless you find out where its source lies,” says Jai Asundi, director of communication and research at a Bengaluru-based think tank Center for Science, Technology and Policy (CSTEP). 

42 million

The government had earmarked $42 million on implementing the National Clean Air Programme between 2018 and 2020. The programme, a framework to reduce air pollution in a time-bound manner, directs CPCB to enhance the monitoring system in India

And even if monitors can roughly point to a source in a particular locality, the data can’t legally be used to shut it down. On the other hand, if this attribution comes through an official monitor, the readings can be used to shut down a particularly heavy polluter, like in the case of the Badarpur Coal Power Station in Delhi. It’s now a site for a new ecological park.  

Singh agrees that monitors, if certified, can be used to fill in the gaps in the official monitoring system. Especially in rural areas, which, on average, have less than 40% of monitors in a state. Until then, he says, CPCB will focus on increasing the number of Continuous Ambient Air Quality Monitoring Stations (CAAQMs) in state capitals and tier I and II cities. 

Their target? 202 CAAQM stations across 114 cities by late 2020. 

Strength in numbers

Even at an accelerated pace of 200-odd stations a year, CPCB still needs to install 20X the number of stations to build up a robust monitoring network for all of India. It has been in talks with the National Physical Laboratory (NPL) to create Indian certification norms, so that monitoring procurement is easily and locally sourced. 

According to information sourced by The Ken, while a proposal to certify low-cost monitors has been in the works, there is no definite time frame that NPL has put out yet. 

“There are minimum three stations in every major city. Why do we need more data? Even if the government were to get more hyperlocal data, it’s not in a position to take quick action. Air pollution numbers have become highly politicised”

Senior Government Scientist

“It will take at least 5-10 years for certification standards for commercial air monitors to come up,” says a government official, who wished not to be identified as he is not authorised to speak with the media. CPCB’s Singh clarifies that NPL is only focused on standards for regulatory equipment.

Certification for low-cost monitors, say experts, is less likely than official equipment as there are no pre-existing standards globally. “You’ll need to have full knowledge of all the failure modes of these sensors. Each sensor is different. So the certification body will have to know how quickly the parts degrade,” says Asundi. 

Air monitoring startups, already in the business of predictions, can’t wait around for this certification to come through. In the last few months, AMBEE has seen its API requests go up by 200-300 every month. Now is hardly the time to slow down. 

Paying attention

Murals depicting Delhi’s poor air quality data and pollutants have sprung up on walls around the capital. To build awareness

Air Veda’s Gupta is also hopeful that state governments, which don’t have extensive budgets to deploy regulatory grade equipment, will lean more heavily on low-cost monitors in the future.

These startups want CPCB’s resistance to low-cost commercial devices to come down, but they are also aware that lowering the bar does not mean compromising on quality. In fact, entrepreneurs say they would welcome a government-approved filter, which will help prevent a sub-par industry of monitors from springing forth. 

With proper certification, low-cost monitors can unlock the potential of hybrid systems within cities, and reach remote areas in states such as Uttarakhand, where a majority of the population still depends on burning firewood—and which has exactly eight CPCB monitoring stations. 

“Imagine the air they’re breathing,” says Anand.

Leave a Comment