Dr. Abir Chakravorty: Assistant Professor at IIT Kharagpur, India
Article Details
Accelerated detection of fruit juice adulteration through UV–vis spectroscopy and data-driven techniques. This study presents a machine learning-assisted UV–Vis spectroscopy framework for detecting and quantifying fruit juice adulteration. Four fruit juices (pomegranate, mango, guava, pineapple) were intentionally spiked with orange juice at concentrations ranging from 5% to 30%. Spectral data were collected in the 300–800 nm range and preprocessed using baseline correction and normalization. Both classification (pure vs adulterated) and regression (percentage estimation) models were developed using an 80:20 train-test split. Multiple algorithms including Random Forest, Gradient Boosting, SVC, KNN, Elastic Net, SVR, and CatBoost were evaluated, with classification models achieving over 90% accuracy and F1-score, and CatBoost performing best in regression with strong predictive performance (R² ≈ 0.80 on test data).
Novelty
The novelty of the work lies in combining UV–Vis spectroscopy with a comparative machine learning framework that addresses both detection and quantification of adulteration in a unified pipeline. The study also evaluates multiple modern ensemble and nonlinear models, particularly highlighting CatBoost for spectral data interpretation. The use of multi-fruit systems and a continuous adulteration gradient (5–30%) adds practical relevance compared to binary or single-fruit studies commonly reported in literature.
Impact
The work has strong implications for rapid food authentication and fraud detection. It demonstrates that high-accuracy adulteration screening can be achieved without traditional chemical or chromatographic methods, significantly reducing analysis time and cost. This can directly benefit food industries, quality assurance laboratories, and regulatory bodies by enabling faster decision-making and scalable testing frameworks for fruit juice authentication.
Originality
The originality is moderate, as UV–Vis spectroscopy and machine learning have been previously applied to food adulteration problems. However, the integration of multiple ML models for both classification and regression, along with systematic benchmarking across algorithms, adds incremental originality. The study is more of a strong methodological integration and optimization effort rather than a completely new analytical concept.
Experimental Rigor
The experimental design is reasonably structured, with controlled adulteration levels, standardized spectral acquisition, preprocessing, and multiple model evaluations. The use of both classification and regression metrics (accuracy, F1-score, R², RMSE, RPD) strengthens the analytical depth. However, the rigor is limited by the absence of external validation datasets, real-world sample testing, and limited adulterant diversity, which may affect generalizability.
Sustainability Impact
The method supports sustainability by reducing reliance on chemical reagents, solvents, and consumables typically used in conventional food testing methods. It also minimizes laboratory waste and enables non-destructive testing. In the long term, such digital spectroscopy-based systems can contribute to more efficient and resource-saving food quality monitoring systems, although the requirement for instrumentation still introduces some environmental and cost overhead.
Applicability
The approach is highly applicable in food quality control environments where rapid screening is needed. It can be deployed in juice manufacturing plants, regulatory inspection systems, and research laboratories. With further development, it could be extended to other food products such as milk, honey, and alcoholic beverages. However, real-world deployment would require model recalibration for different processing conditions, seasonal variation in raw materials, and broader adulterant types to ensure robustness.
Research Portfolio
Abir Chakravorty is an Assistant Professor at IIT Kharagpur working at the intersection of food engineering, chemical engineering, and intelligent automation systems. His academic work emphasizes building practical technologies for food safety, quality monitoring, and process optimization using robotics, AI, and advanced sensing methods. He actively contributes to both research and teaching in food process engineering and industrial food systems.
Online Profile
Education
He completed his undergraduate studies in Chemical Engineering from West Bengal University of Technology with strong academic performance, followed by a Master’s degree from the University of Calcutta. He later earned his Ph.D. from IIT Kharagpur, where he specialized in mass transfer phenomena, reaction engineering, and multiphase flow systems. His doctoral training provided a strong foundation for applied research in food and process engineering.
Research Focus
His research primarily focuses on food process engineering, automation in food manufacturing, and AI-based quality assessment systems. He works extensively on biosensors, carbon quantum dots, and smart detection systems for food safety and adulteration analysis. In addition, he explores advanced fluid dynamics, membrane processes, and process intensification techniques such as pulsatile and multiphase flow systems.
Experience
He has been serving as Assistant Professor at IIT Kharagpur since 2022, contributing to teaching, research, and institutional responsibilities. Prior to this, he worked as Senior Project Associate at the West Bengal Pollution Control Board and conducted research in chemical engineering domains, including perovskite solar cell studies. His early career includes industry exposure as a Graduate Engineer Trainee at Asianol Lubricants and training at Himadri Chemicals & Industries Limited.
Research Timeline & Activities
His research journey began with chemical engineering-focused doctoral work from 2014 to 2021, emphasizing multiphase flow and transport phenomena. Post-PhD, his work shifted toward applied food engineering, robotics, and AI-driven systems. From 2022 onward, he has expanded into funded projects involving robotic food processing systems, AI-based food adulteration detection, DRDO-supported space nutrition research, and sensor-based quality assessment technologies.
Awards & Honors
He has received multiple recognitions including the Best Poster Award at CHEMCON 2017 and a Novel Concept Award in 2023 for research on AI and biosensors in food shelf-life prediction. He also received MHRD fellowship support during his PhD studies. Several of his students have also earned awards at national-level seminars under his supervision, reflecting his mentorship impact.
Strengths for the Young Researcher Award
Strong interdisciplinary integration
The researcher demonstrates a clear ability to integrate food engineering, chemical engineering, and artificial intelligence, which is a highly valued trait for early-career recognition. His work spans spectroscopy, machine learning, biosensors, and process engineering, showing intellectual versatility rather than narrow specialization.
High relevance to real-world societal problems
A major strength is the focus on food safety, adulteration detection, and quality assurance, which directly connects academic research to public health and industrial needs. This applied orientation increases the societal value and practical impact of his research portfolio.
Rapid growth in research impact
Bibliometric indicators such as citation growth concentrated after 2021, along with an increasing h-index and active publication trajectory, indicate a fast-rising research profile. This pattern is often favored in young researcher awards because it shows momentum and future potential.
Strong emphasis on AI-enabled engineering solutions
A key strength is the consistent use of data-driven and AI-based methods in engineering systems, particularly in food authentication and process monitoring. This positions the researcher at the intersection of traditional chemical engineering and modern intelligent systems, aligning with current global research trends.
Effective academic leadership and mentorship potential
The researcher has demonstrated early academic leadership through teaching, student mentorship, and supervised award-winning student work. Combined with institutional responsibilities at a premier institute like IIT Kharagpur, this reflects strong potential for future research group development and academic contribution.