A²I : Scrutinising different types of green-groceries through machine vision
A holistic approach to improve food supply chain efficiency
Food, shelter and clothing are the most basic and essential needs of mankind. The challenge of supplying nutritiously fresh, affordable and processed foods to 9 billion people in 2050 will require a 70 percent increase in food production which can be met by strategizing to maximize the efficiency of available resources. Hence, Food processing is one of the major manufacturing sectors in the world. Food processing mainly involves transforming raw food ingredients by using a set of methods and techniques to produce marketable food products that can be easily prepared and served to the consumer at affordable prices.
For food manufacturers, Artificial Intelligence(AI) is being used to improve both their production processes and their products. Apparently, it might not seem to be the case but, food processing is actually dominated by big data problems. For problems in consumer desire, ingredient sorting, and recipe development — AI is providing the most workable alternative to human expertise.
Food processing often comprises sorting a large quantity of feedstock and careful inspection of the final product. Among manufacturing operations, one challenge that is relatively unique to food processing plants is that the feedstock is often not uniform. To know if, say, a potato or an apple is usable requires an assessment of its size, shape, colour, and marks. At the turn of the 20th century, according to TOMRA, a leader in food sorting technology, 90 percent of all food sorting was done by professional inspectors who were capable of making these highly complex assessments at a reasonable speed.
A comprehensive approach for improving food supply chain efficiency and developing sustainable agrifood systems to the ever-changing world is needed. In order to achieve this, automatic systems are now being developed to collect hundreds of pieces of data on a single produce and rapidly make an assessment about it within a few seconds.
The automated ripeness detection system developed by computer vision scientists at Amazon consists of a conveyor belt that transports the food in containers to a particular sensor.The sorting machines uses cameras, Near Infra Red (NIR) spectroscopy, X-rays, and lasers to gather and process all that data from hundreds of individual ingredients as they rapidly move across a conveyor belt. By giving new product variants daily, the machine learning algorithm learns about how to differentiate about between good and bad produce. The computer gradually understands the quality standards. The fresh produce is divided into four categories: “OK”, “Damaged”, “Badly damaged” and “Expired”. Containers of the different ripeness categories are randomly offered to the machine to prevent the machine from learning a pattern and allows it to perform proper checks.
Machine learning creates the capability to sort foods by sending the exact same discernment every single time. The problem of good produce being mistakenly thrown away, or expired product being delivered to the consumer will be a thing of the past as these systems can produce higher yield with optimised quality and also bring down labour costs significantly.
Due to the growing interest in reducing the subjectivity using unique and non-contact measurements,a detailed experimentation is carried out using various machine learning and computer vision techniques. Computer vision systems is focused on defining new ways for the evaluation of colour and shape parameters. The colour receives special interest because it is an important sensory attribute providing necessary quality information for human perception.In the fruit inspection industry, the support vector machine (SVM), k-nearest neighbour (KNN), artificial neural network (ANN), and decision tree (DT) pattern classification methods are the most commonly used (Arabasadi et al., 2013; Vithu and Moses, 2016).
Another classic example of advancement of AI in agrifood sorting industry is where a Brazilian research group worked vigorously on developing an application that enables papaya growers to send the ripest papayas to local markets and saving less ripe papayas for export. Hence, reducing food waste and appreciably adding value to the product.
Hence, AI is not a futuristic concept — it can be leveraged today to make operations more productive, efficient and profitable. With the speed of the marketplace and the rate of exponential change in the resource-constrained world, exciting insights in the food industry will be possible in the near future.
R&D at Arnekt
Arnekt helps in turning the vision into action. Arnekt envisions development of AI-powered applications and web services using Cognitive Intelligence. Arnekt mainly focuses on building data-driven algorithms to improve logistics for businesses without compromising on the performance which helps in making greater impact as a leader.This helps in giving a competitive edge and creating business value.
