Currently, SLCM is conducting trials for wheat and Chana. It intends to cover rice, guar, moong, tur and other pulses.
Sohan Lal Commodity Management Pvt Ltd (SLCM), a post-harvest crop management solutions provider in India has developed an AI-based mobile app to provide a quality assessment of a crop within a minute with up to 90 per cent accuracy.
The app aims at improving the fidelity of commodity quality checks and bringing transparency by reducing turnaround time for quality check. The app will be based on Machine Learning with Python. Once the image of the crop is uploaded on the app, the system will read the image and create a database based on the physical parameters such as height, length, grid, colour, specific pattern.
While explaining the use of the AI-based app, Sandeep Sabharwal, CEO, SLCM Group said, “In the AI-based mobile app, the trader or farmer would just need to scan the seeds through SLCM mobile app. It can check multiple quality parameters for a variety of agri produces within a minute with 85-90 per cent accuracy as compared with lab test. It compares with pre-fed data in our back-end system, which also keeps updating itself with machine learning. This is the on-ground utilisation of the technology for agri trade.”
Sabharwal also added that with the recent advancement in AI and Machine learning technologies, SLCM does foresee business cases to reduce TAT (Turnaround Time) in doing quality checks – then and there at “Mandis” or warehouse using our AI-enabled mobile app instead of sampling and sending it to distant labs. On predefined quality parameters we want to achieve the same level of accuracy as in our current lab-based reports. Such type of accuracy will nudge further business decisions almost in run time. Accuracy is going to be a feature that keeps on going better as the machine learning model of the app keeps maturing with the training dataset. Improve the fidelity of commodity quality checks – transparency will be enforced by linking the quality check process with our AI-enabled mobile app.”
