The inclusion of advanced technologies such as AI and machine learning in offline retail ensures seamless operations and minimises operational costs in offline retail. To redefine offline retail in a new avatar, BigBasket recently has acquired the enterprise business unit of Kerala-based deep tech startup Agrima Infotech to power BigBasket’s self-checkout offline store Fresho with its computer vision technology platform – Psyight.
Agrima Infotech’s Psyight is the world’s first food-focused recognition and Information Orchestration Platform that will aid BigBasket to identify all Indian fruits and vegetables uniquely from an image without using barcodes. Detecting raw food items like fruits and vegetables uniquely from an image is a complex machine vision problem since the appearance of those items changes drastically according to the location of its origin and seasons. Psyight’s AI recognition feature is versatile and can be integrated into any food system having a camera. Through this, AI recognition platform thousands of SKUs across seasons and various locations can be scanned to achieve 100 percent accuracy.
‘‘
A CIPHET study finds that the post-harvest waste and loss in the fruits and vegetable segment in India stands at 40,811 crores.
How AI recognition will work magic for offline retail?
Reduce inventory losses and wastage: Fresh fruits and vegetables in transit from the farmer’s place to the retail store undergo deterioration in quality such as green leafy vegetables losing water or apples shrinking in size, gradually moving closer to the expiry date. The change in the size of fruits and vegetables goes unnoticed by the naked eye. This prevents the retail owners from selling first the inventory nearing the expiry date. AI recognition will analyse the shape of fruits and vegetables that will help retailers or grocers to sell stocks of near expiry first and keep them for little time, thus preventing losses and wastage. It also helps supermarkets to plan inventory with the help of an algorithm that analyses previous demands and sale trends.
Increase profits and sales through an optimised planogram: The AI product recognition can identify which items are missing from the shelf to remind the store staff to restock products immediately thus following planogram compliance of products on the shelf. According to a study by National Association for Retail Merchandising Services (NARMS) when an optimized planogram is 100% matched, sales will be increased by 7.8% and profit by 8.1.
Increases efficiency at the checkout: Long billing queues are a disappointment for a majority of retail shoppers in India. Automatic product recognition ensures an efficient checkout process. Until now cashiers have had to type in the barcode for all fruits and vegetables for billing but with automatic product recognition, this will no longer be necessary.
Accurate stocks: Had there been barcodes, there exists a probability that an organically grown potato is accidentally scanned as a normal one impacting the accuracy of the actual count of organic potatoes. But, AI recognition on scanning reflects this in the store data. This also solves store-level ordering conventionally done using pen and paper, which is considered to be the biggest pain point in the world of fresh food.
Can grocery stores in India adopt AI recognition?
A CIPHET study finds that the post-harvest waste and loss in the fruits and vegetable segment in India stands at 40,811 crores. 5 to 16 percent of perishables like fruits and vegetables are ruined in transit or on farms therefore the appetite to adopt AI recognition systems or platforms in grocery retail exists. But lack of internal expertise and fear of complexity hinders them to adopt the same. For AI to be used in grocery retail, somebody has to train an AI model that knows what and how to do what the customer wants. The AI models require to be retrained constantly once integrated with the store. Retraining is necessary since AI models tend to lose accuracy and performance over time.
