Machine vision utilising artificial intelligence

IVS has witnessed an increasing number of customers investing in machine vision systems to reduce recall figures and safeguard their revenue, reputation and brand worth.

In June 2019 alone, well-recognised UK brands such as Tesco, Aldi, Dunelm and Asda have recalled items such as sushi salmon, cotton towels, tomato seasoning, and liver and bacon mash. Previously, Samsung was forced to launch a global recall for its flagship phone, the Galaxy Note Seven, after batteries kept exploding, while faulty airbags from car component manufacturer Takata led to the largest recall in automobile history. Moreover, in 2015, General Motors had to recall millions of cars because of faulty ignition switches that led to fatalities, lawsuits, and a $900m criminal penalty. 

“The impact of issuing a recall can be devastating in terms of cost, reputation and, eventually, market share. Over the past year, unparallelled levels of developments have occurred in artificial intelligence (AI), big data, 3D imaging, and robotic process automation – none more so than on the factory floor – and this is crucial in reducing the risk of error and subsequent product recalls,” said Earl Yardley, Director at Industrial Vision Systems.

Yardley believes that advances in thermal imaging, 3D imaging, and hyperspectral imaging are amongst the developments which will play a vital role in assisting manufacturers in reducing the number of product recalls.

Yardley concluded: “Machine vision is used for final quality control inspection, but new markets are opening up with the advent of 3D sensors and deep learning utilising artificial intelligence. Deep learning within machine vision can detect defective parts traditional vision inspection systems cannot see. 

“Typical applications in hyperspectral imaging include plastic detection in meat production, detection of different recyclable materials and blister pack quality control. While development in automated thermal imaging inspection allows manufacturers to spot problems which can’t be seen by eye or standard camera systems and provides non-contact precision temperature measurement and non-destructive testing.”