Though the agriculture industry in India is one of the largest, it is often unorganised especially in the post-harvesting supply chain where it witnesses 30-40% food losses due to the major gap between supply and demand, and inefficient logistics. These losses result in increased prices to the customers with farmers getting little share of the price paid by the consumers.
Another major challenge in the value chain is the time it takes to reach the market, which is usually 24-30 hours, resulting in wastage in many cases.
Karan Hon, who comes from the farming family along with Puneet Sethi recognised these challenges very well and after a two-decades-long journey in the corporate world, decided to set up Farmpal, a Pune-based startup that is on a mission to organise the post-harvest supply chain so that farmers can have better access to alternate markets. It also promises to provide fair prices for their produce and reduce inefficiencies across the supply chain so that all stakeholders are benefitted.
To achieve their goals, Farmpal is extensively using artificial intelligence and other technologies to connect farmers directly with end consumers such as local kiranas, supermarkets, sabjiwala to local societies and more.
Analytics India Magazine got in touch with Puneet Sethi to understand how AI comes into the picture at Farmpal which provides a mobile app, connecting farmers and customers at the front-end. “This is backed by a powerful ERP solution at the backend that helps us streamline inventory management, order cycles, supply-demand forecasts and analytics for all other functional areas. One of our key goals is to reduce wastage across the supply chain. For that, a close match of supply-demand is essential which is where AI plays a key role,” shares Sethi.
Farmpal’s AI-Based Solution Has Been Able To Predict The Demand With 95% Accuracy
One of the ways to deal with wastage in the agriculture industry is to predict the consumer demand and meet the supply accordingly so that there is minimum wastage across the value chain. The AI comes into the picture to predict the demand, where Farmpal has designed an artificial intelligence solution to predict demand at 95% accuracy and keeping the wastage below 5%.
As Sethi shares, they are leveraging Oracle ERP historical data of supply-demand and placing a layer of AI over ERP to curate historical data. “AI makes sense out of available data to understand variation in demand considering peak and off-peak days in a week for each customer segment and seasonal variation for each SKU considering festive seasons,” he said.
AI modules have helped them understand the weekly demand for more than 50 SKUs where they are able to inform the demand to farmers a week before, and farmers can plan on harvesting based on this demand to avoid over or under supply, resulting in minimum wastage.
The fact that Farmpal has historical data of over two years and an extensive study on the post-harvest trends and patterns both in terms of production and consumption, helping them to accurately design AI solutions. Farmpal leverages a lot of this data and publicly available data to predict and forecast demand and supply patterns.
“As our volumes grow and historical data grows with it, we are able to do a deeper dive into the trends and patterns such as production and demand for specific seasons, months and even specific days of the week. All this data is fed to the AI engine and algos, along with certain other conditions to forecast the demand and then enable matching supply through an integrated farmer database where we store farmer information such as profile, land holding, type of produce, harvest cycle, capacity etc.,” explained Sethi.
The tech stack at Farmpal is a homegrown software coupled with Oracle ERP to enable their tech functions, including AI. “The Oracle suite is pretty comprehensive and provides a host of AI capabilities, specifically AI for the supply chain functions,” shared Sethi.
Other Areas Where Farmpal Is Using AI
Currently, the startup is using AI for the supply-demand cycle for different times of the year, week etc., they are also exploring AI for more efficient route planning and route optimisation to better manage the logistics.
“In future, we will use AI for customer data to model, and then leverage information about purchase patterns, including their basket size allocation, so our pricing is more in sync and to enable the use of customer promotions to boost sales and also build customer loyalty,” he added.
Farmpal solutions are so far deployed for their own internal working and tech enablement to support operations across Maharashtra.
Challenges In Developing AI Algorithms For Agriculture Industry
While the solutions seem promising, the Farmpal team has to face challenges working in the agriculture sector. The primary challenge, as Sethi shares, is the lack of historical data that is very specific to the post-harvest cycle, in terms of the supply chain and the supply-demand forecasts.
“Additionally, the Agri supply chain still works primarily on old antiquated mechanisms where both farmers and high volume purchasers go to the local APMC or mandis. So the initial challenge was to have available meaningful data for AI algos,” he said.
Secondly, especially with fresh produce, the perishability factor plays an important role that has to be factored in.
Ensuring Direct Supply To Societies In The Covid-19 Times
While Covid-19 caused disruption in Farmpal’s ground operations and they scaled down due to the lockdown and non-availability of manpower and logistics, being essential services, they were able to get back necessary permissions. “Keeping the challenges in mind, we have narrowed down our focus to high customer density areas both to reduce manpower needs in our collection and distribution centres as well as to optimise transportation,” he said.
Having said that, they have piloted a new segment during these Covid-19 times, that is, ensuring direct supply to societies or a collection of customers in a local area through a “business owner” model.
Farmpal continues to focus on increasing the footprint both on the supply side, in terms of farmer tie-ups across Maharashtra and in other States; and on the demand side by retailer customer acquisition in Pune, and then Navi Mumbai, Mumbai and cities in Gujarat. The startup is also looking at increasing supplies to supermarkets in other States.
They also plan to improve upon the use of machine learning and artificial intelligence to better predict the demand and map the supply to continue ensuring minimum wastage. He also added that on the operations side, they plan to use IoT, RFID technologies, route planning solutions, advanced analytics in the coming future.
“On the farmer front, through our platform, as we onboard and connect with farmers across India, we want to be able to better assist farmers by providing better access to the Agri ecosystem,” said Sethi on a concluding note.
Join Our Telegram Group. Be part of an engaging online community. Join Here.
Subscribe to our NewsletterGet the latest updates and relevant offers by sharing your email.
Srishti currently works as Associate Editor at Analytics India Magazine. When not covering the analytics news, editing and writing articles, she could be found reading or capturing thoughts into pictures.