Researchers at the National Institute of Technology Rourkela developed a technique to quickly detect and measure adulteration in spices and other food products. The Institute has also secured a patent for the technology. Addressing the global food safety challenge, the developed technology combines Fourier Transform Infrared (FTIR) spectroscopy with advanced machine learning models to deliver accurate results.
In the Indian context, food and spice adulteration poses serious health and economic risks. It is often driven by cost-cutting practices and inadequate food safety standard checks.
Traditional food adulteration detection methods such as chromatography, or molecular techniques, are resource-intensive and take a large amount of time to deliver results, making them less suitable for rapid, routine testing.
To address these limitations, the system developed by NIT Rourkela researchers provides a rapid, non-destructive, and cost-effective alternative suitable for real-time deployment in quality control laboratories and industrial processing units.
FTIR spectroscopy is a technique used to identify organic, and some inorganic materials by measuring how they absorb infrared light. During food checks, the developed system collects these patterns and processes them using machine learning models. These models look at complex, non-linear patterns in the sample to detect abnormalities and give accurate results on adulteration levels.
Unlike conventional methods that only show whether the food product is adulterated or not, the developed technology measures the level of adulteration of food within seconds. This feature is critical for food processing industries and regulatory bodies that require precise measurements to ensure compliance and maintain product quality.
Published in the prestigious Food Chemistry journal, the research has been conducted by Prof. Sushil Kumar Singh (Assistant Professor), Late Prof. Poonam Singha, and M. Tech. graduate Mr. Rishabh Goyal from the Department of Food Process Engineering, NIT Rourkela.
Study on detecting coriander powder adulteration
Beyond the initial validation of the developed system, the research team also addressed a common and concerning adulteration practice, the mixing of sawdust in coriander powder. By applying machine learning models integrated with FTIR spectroscopy, the team developed a framework capable of analysing and detecting adulteration with around 92% accuracy.
According to Prof Sushil Kumar Singh, “Any food company that processes spice at any stage, from raw materials to finished products, requires rapid adulteration detection. Our developed system can seamlessly integrate into their existing quality control workflows and allow real time decision making, which is highly suitable for routine screening. With its scalability and cost-effectiveness, the system has strong potential for adoption by both large industries and SMEs.”
The system developed by the NIT Rourkela team reduces production delays by eliminating the need for complex sample preparation and longer analysis time. This cost advantage is relevant for price-sensitive markets, such as India, where large scale screening and affordable food safety solutions are critically needed.
As the next step, the research team aims to collaborate with industry partners for conducting pilot-scale studies and validating the system under real-world conditions. Additionally, they plan to conduct experiments under different conditions to extend its detection capability beyond spices.