Murugan Chidhambaram is the Head of Digital Transformation at Aquaconnect, Chennai. As such, Murugan leads a team of engineers and IT professionals passionate about creating and delivering solutions that leverage machine learning, Geo Information Systems, Cloud Computing, Data Science, and Visualization to transform the Aquaculture industry. 

He also manages the IT business by leading the monitoring and reporting of program-related trends and capability developments. 

INDIAai interviewed Murgan Chidambaram to get his perspective on AI.

How does Aquaconnect leverage AI and satellite remote sensing to improve the seafood value chain? 

Aquaconnect leverages Artificial Intelligence and satellite remote sensing to enhance transparency and efficiency in the aquaculture value chain. Using geospatial data powered by deep learning models, we can accurately define pond boundaries, distinguish between fish and shrimp ponds, and predict culture days. This intelligence also helps forecast the demand for farm inputs and the supply of harvest produce, scaling from individual ponds to village and national levels. 

Our unique technology stack combines satellite remote sensing “eyes-in-the-sky” and field intelligence "boots-on-the-ground”. It uses deep learning algorithms to identify and map fish and shrimp ponds. Remarkably, the AI models we have trained now identify errors in our initial training data, showcasing their advanced capabilities. We subsequently developed additional AI models based on various parameters to differentiate between shrimp and fishponds. The next critical step was predicting the Days of Culture (DOC), which helps us assess the pond's status—whether active or inactive, in preparation, stocking, grow-out, or ready for harvest. This hierarchical crop review is conducted near-real time without physically visiting the farm. Understanding DOC is essential for driving various use cases and optimizing aquaculture practices. 

This integrated approach enables efficient market linkages between stakeholders, aiding better access to farm inputs and connecting post-harvest markets to sell harvest produce. 

How did Aquaconnect decide to integrate AI into its operations, and what were the initial challenges you faced? 

As an aquaculture tech startup, we initially launched a mobile app as a farm advisory tool to capture production data, guide farmers in their day-to-day culture operations, and connect them to procure farm inputs and sell their harvest. Although this app helped us understand the correlation between different parameters, feeding, and growth patterns, the challenges we encountered were adoption among farmers and getting pond data at required intervals. We understood the steep learning curve and practical difficulties. Then, we brought a chat-based interface to improve the adoption. But it begged the question- how scalable are these? How can we capture intelligence from the nook and corner of India and the rest of the world? 

The ideal answer is that every farmer should have a smartphone, better connectivity, and mobile app awareness. Finally, he must keep on entering the data daily, no matter what. That is when we started looking at innovative solutions and discovered how satellite remote sensing could be a game-changer for aquaculture. By focusing on scalable solutions that address critical challenges in the sector, we shifted our attention to harnessing the power of satellite remote sensing and artificial intelligence (AI). Using satellite remote sensing, we gained “eyes in the sky”,; breaking geographic limitations and enabling us to scale from individual ponds to the national level without physical constraints.  

Can you explain how satellite data and AI have led to actionable insights and improved aquaculture practices? 

Aquaconnect collects satellite and ground data intelligence, then processed through machine learning algorithms. This process helps predict three critical factors: what to sell, where to sell, and how to maintain optimal inventory levels. This predictive capability enables our Aqua Partners to forecast farm input demand in their vicinity and identify potential customer bases to upscale their business. 

One remarkable capability of our satellite remote sensing technology is its ability to identify the precise age of a shrimp crop. Whether it's a 10-day-old crop, a 30-day-old crop, or a 90-day-old crop, we can determine its developmental stage. This capability has profound implications for assessing regional demand for fish and shrimp feed products and predicting harvest supply. By predicting harvest, our models can solve critical sourcing challenges for seafood buyers, enabling them to plan their procurement more effectively. 

In what ways is AI helping Aquaconnect promote sustainable aquaculture practices? 

The sector is in the early stages of adopting AI. If we look at the ground reality, just tech solutions cannot drive sustainability; instead, we need to embrace one more AI (“Ancestral Intelligence”), and both must complement each other. Artificial Intelligence can be an enabler, and we educate and promote sustainable practices among farmers through our Aqua Officer farm visits and “Farmer Talks” awareness programs. Looking at the larger picture, the models have tremendous potential--One such potential use case is that the intelligence we capture can be leveraged to build traceability of harvest produce. Another one we are exploring is quantifying the carbon emissions, measuring the environmental footprint and strategizing countermeasures. 

What future developments do you foresee in applying AI in aquaculture, and how is Aquaconnect preparing for these advancements? 

Over the past two years, we have all felt the ripple effects of AI in our daily lives. Given the pace of evolving AI models, it will have a phenomenal impact on the sector beyond mere data analysis. Imagine a friendly farm companion integrated with deep learning, capable of real-time monitoring, providing actionable insights into water quality, optimizing feed intake and FCR, monitoring animal health and recommending appropriate healthcare solutions, and offering guidance to farmers on the day-to-day challenges of aquaculture, uncovering patterns and insights from historical data previously unseen. They cannot just impact production; they can broadly affect the post-harvest chain and policy-making, support strategic decisions to improve the sector’s output and contribute to the nation’s food production. 

Looking at the bigger picture, we envision leveraging AI to bring predictability and efficiency to make macro-level decisions supporting value chain players in accessing inputs and post-harvest markets. We are experimenting potential AI use cases and constantly evolving to integrate our solutions that can address the sector’s challenges such as lack of traceability, and measuring carbon quantification etc.,  

What advice would you give aspiring technologists and entrepreneurs looking to make a difference in the agri-tech sectors? 

The biggest challenge in AgTech is not building tech but making it work for the people in the intended way. The secret sauce is that you must innovate the business model, create awareness, and pricing and last-mile distribution/connectivity. I highly stress making your solution affordable, scalable, and accessible to the stakeholders. You must meet the practical needs and offer a considerable delta in value realization to ensure the solution succeeds. We must consider upgrading beyond technology-centric business models to deliver, support, and finance these solutions, making them economically viable and user-friendly.