‘Can adapt our covid model to any infectious disease’

Shipra Misra is chief executive and managing director of Delhi Research Implementation and Innovation (DRIIV), the umbrella body for the Delhi NCR Science and Technology (S&T) cluster. In an interview, she talks about the progress the foundation is making with Project SAMEER to control air pollution in Delhi, the move to introduce solar-powered interoperable electric vehicle (EV) charging stations, and the use of artificial intelligence (AI) in healthcare, among other things. Edited excerpts:

How effective are these S&T clusters proving to be? What are you focusing on?

India has six S&T clusters—Delhi, Hyderabad, Bhubaneshwar, Jodhpur, Pune, and Bengaluru, and each is named after the city they are based in. But their activities are not confined to those cities. All these clusters are relatively young, and our projects are aligned with the national missions of the country. Our partners too play an important role in taking innovation from the lab to the market.

How much progress has the Delhi S&T made on air pollution with Project SAMEER?

We have to take a holistic approach to overall pollution. Hence, we have been collaborating with bodies such as the Gurugram, Metropolitan Development Authority (GMDA), the Municipal Corporation of Delhi (MCD), and Haryana pollution control board in areas that cover waste management, water security, air pollution, landfill, traffic management, healthcare, etc. We picked air pollution as the first use case. Project SAMEER (Solutions for Air-Pollution Mitigation through Engagement, Engineering, and Research) is employing a three-pronged strategy where we are focusing on awareness and community engagement, deploying science and technology solutions, and mobilizing the industry, with the help of local government authorities, IIT Delhi researchers, tech startups, corporates, and NGOs.

What’s the approach, and how are you measuring its efficacy?

As part of our awareness and community engagement, we have targeted three communities. The first is farmers in Punjab and Haryana, who are being made aware about the impact of stubble burning on their lungs. 

We are doing this in partnership with NGOs such as the Lung Care Foundation and the India Paryavaran Sahayak (IPS) Foundation. We are also targeting school and college students to develop a scientific temper in them and help teachers develop pedagogical tools that will make these concepts easy to understand. Besides, there are about 12 startups that have volunteered and offered their solutions free of cost to be deployed across sites in Delhi and Gurugram. Some deployments have already happened, some will be deployed over the next few weeks. Simultaneously, a team of researchers from IIT-Delhi is monitoring the results of these projects, following which a detailed report will be submitted to the local authorities, after which we will figure out the next course of action.

Give us examples of some deployments?

It’s a mix of solutions. Some are focused on monitoring PM (particulate matter) levels. Others are using low-cost sensors and low-cost monitors in strategic areas to reduce implementation costs. We are also using ionization-based misting (to help oxidize chemical contaminants to a simpler form with no harmful compounds generated) solutions. The machines, which resemble a refrigerator, can be deployed in industrial areas during high pollution days. Thus, you will not need to shut down these areas that typically result in loss of livelihood.

You are dealing with a lot of data. How is AI/ML helping you in this task?

Let’s talk about one use case in healthcare. We developed this platform during covid-19. So, it’s an MDM (mobile device management) compliant platform. And a number of models are based on that. 

This model was trained on genomes, and it would identify genome sequencing, and whenever there was a change, or spike, in the genome sequencing, the model would catch that. 

Typically, these spikes occur up to two months ahead of the infection reaching pandemic levels. What if you could predict this at least two months in advance? This would give policy makers enough lead time to plan their action. 

This model, which was developed in conjunction with ICMR (Indian Council of Medical Research), This AI/ML strain prediction model can be adapted to any infectious disease and can prove to be an effective tool for policy makers and healthcare agencies to manage community health. The other is using sensors or IOT (internet of things)along with AI/ML to monitor and measure air quality in real time.

Tell us about some of the research in the area of EV batteries and charging stations.

Researchers have developed charging solutions suited to Indian conditions, and those that can tap power from the grid. The challenge is that it is not helpful if the grid itself produces electricity from coal. So, the idea is to make this charging completely off-grid. In this context, researchers are trying to integrate solar panels in the charging stations and add battery packs to make them work even at night. Such charging stations have been set up within the IIT-Delhi campus and a few other places in Delhi. We are now planning deployment of a solar panel-enabled charging station with battery packs, following which even the ones within IIT-Delhi campus and other places will become solar-powered.

How do you plan to scale up these pilots?

The most efficient way of scaling it up would be to develop a business model around it. We are trying to develop a low-cost charging station where the shop owner can retain a portion of the revenue that is generated from charging after paying the rent and electricity (when taken from the grid). If the charging is done with solar power, the profit could be higher but this model that is being tested is not very popular because of the high capex cost. So, we are working on a separate project to build interoperable charging stations. If this made available, it will make it very easy for anybody to set up these chargers, and low-cost business models can be developed. Also, these charging stations would have to be portable, which will allow even kirana (mom and pop) shop owners to set it up. It may six months on the higher side to integrate the charging stations with solar panels and battery packs. Interoperability will take around 18 months to two years.

What research is being done in the area of solid waste?

We already have a plant at the New Jafrabad Technology Park in East Delhi that is currently using plasma pyrolysis (producing high temperatures using a plasma torch in an oxygen-starved environment) to convert municipal solid waste into gases. Those gases are then cleaned and released back into the atmosphere. The next stage is to extract hydrogen out of these very gases and bottle them and use them. But currently this plasma pyrolysis plant runs on electricity. The plan is to convert it into a solar-powered plant so that it becomes self-sustainable. We are also exploring the setting up of such decentralized units near landfill sites too.

What about the focus on green hydrogen, given that we already have some hydrogen buses on the road?

These buses are currently using a mix of hydrogen and other gases, which is also fine. However, in terms of pure green hydrogen, the technologies are still cost prohibitive. Even as some big companies such as Reliance Industries are doing their own R&D to lower costs, nobody has been able to bring down the cost levels to a point when it becomes commercially viable. Researchers, meanwhile, are experimenting with other ways to produce hydrogen other than hydrolysis (to extract hydrogen from water). One such experiment is being done in Mohali where sunlight is being used as a catalyst (producing hydrogen by splitting water using photocatalysts and solar power). If this works, it would make the production of green hydrogen very cost effective.