As a Masters student under Prof. Vikram Pudi, Sri Teja Allaparthi has been working on detection of misinformation in social media. But not many know that apart from developing a deep learning model that tags news articles in media based on classes such as hatred, bias, satire, and so on, he is also the technical brain behind INDRA Water, a unique Mumbai-based startup that is into waste water treatment. Read on.
It started as a casual collaboration during Sri Teja’s 3rd year at IIIT-H, in lieu of a formal industry internship and now he’s part of the team behind a unique method of treating waste water. An idea that germinated in Amrit Om Nayak and Krunal Patel’s Master’s program at the University of Washington led to the building of a prototype and its eventual commercialization once they returned to India. From a team of two, they are now five with the coming on board of Aditya Vemuganti, Abhijit VVR and Sri Teja. “I knew the founders of the startup through my brother who is their friend. They’re mechanical engineers and I’m the only tech guy on the team,” says Sri Teja, speaking of how the alliance came about. Incubated at BioRiiDL, India’s first bio-incubator located at Mumbai, INDRA Water got an initial grant of 10 Lakhs from DST (Department of Science and Technology), through NIDHI PRAYAS program facilitated by IITB. What makes this startup unique is that unlike typical chemical filtration processes followed by most sewage treatment plants, it uses the technique of electro-coagulation, or the passing of electricity to break down the chemical composition of industrial effluents and sewage water from residential and commercial enterprises. When waste water is treated in this manner, about 95% of it is fit for all kinds of household purposes except drinking. “To make this water potable, we need to add extra modules and the current market is not ready for it yet,” explains Sri Teja. Another plus is that their recycling machines occupy upto 40% lesser space than traditional ones, making them a smarter choice in space-starved cities.
The water treatment machines typically require manual intervention in the form of maintenance, checks on daily functioning and other logistics. Being a fledgling startup lacking immediate economies of scale, a decision was taken to automate all the manual processes. “We collect 106 parameters every second from our machine(s). The parameters give us information about the different parts of the machine, as well as quality of input and output water. We are constantly monitoring these values. If any module in the machine starts to malfunction, the quality of the treated water is affected,” says Sri Teja. Appropriate hardware in the form of smart sensors automatically detect malfunctioning in any of the modules when the quality of the output water does not match the predicted quality. “In this way, by monitoring in real-time, we reduce operational costs, costs of electricity as well as eliminate manual costs of tinkering with each module,” says Sri Teja.
What makes the team’s systems even smarter is the way the entire water treatment process is managed. Based on the quality of water that goes into the machine, certain parameters need to be tweaked in order to treat it effectively. For example, the manner in which industrial effluents from the textile and dyeing industry are treated is different from the way sewage water from gated communities is treated. Machine learning algorithms come into play here as the machine learns about the the type of water being treated and fine tunes parameters automatically. An analytics dashboard on their website provides real-time feedback to the consumers and this is currently being offered as an optional add-on feature by the team. Acknowledging the solid base in Data Science and Machine Learning acquired while studying at IIIT-H, Sri Teja says, “We’re using Python frameworks to build machine learning algorithms to get the final analytics and I’m in charge of both the automation platform, Indra Smart as well as the analytics platform known as Indra Spectrum.”
Speaking on behalf of their team about their upcoming goals and milestones, Sri Teja says that he is first looking forward to joining them formally as CTO of the startup upon completion of his Master’s degree. With the startup being selected by the Research and Innovation Circle of Hyderabad (RICH) to find solutions to the increasing water woes in conjunction with the Government of Telangana, setting up of a Hyderabad-based office for the technical team is also on the anvil. “When I joined IIIT-H, I had already heard that the research labs here are comparable to the best in the country. Initially, I wanted to become an expert in ML and go on to complete a PhD. But after coming in close contact with the startup community, I feel this is more practical, applied and challenging,” says Sri Teja. Early this year, with the Lodha Developers shortlisted INDRA Water for building portable water treatment systems at their upcoming Palava Smart City near Mumbai, and interest evinced by some pharma and textile companies in Hyderabad, exciting times are ahead for this young startup.