Amazon Web Services has launched a vision service for inspecting products on manufacturing lines.
Amazon Lookout for Vision is a cloud service, providing a machine learning model that customers can train with their images – as few as 30 baseline images, according to the company – to spot production defects.
The service has no up-front commitment or minimum fee. Rather, customers pay by the hour for using it to train the model.
The machine learning technique Amazon uses is called few-shot learning, where the model is trained to classify new data with only a few samples.
The Lookout for Vision service joins Amazon Lookout for Equipment, Amazon Monitron – both of which are designed to monitor abnormal behaviour in industrial equipment – and AWS Panorama offerings for what it describes as a 'suite of cloud-to-edge industrial machine learning services'.
The service is aiming to open up machine learning technology to many more manufacturing plants by removing some of the barriers involved in developing, training, deploying, monitoring and fine tuning computer vision models.
Customers send camera images to Lookout for Vision in real-time to identify anomalies. It then reports images that differ from the baseline via the service dashboard.
Amazon says Lookout for Vision is sophisticated enough to maintain accuracy with variances in camera angle, pose, and lighting arising from changes in work environments. Customers also have the ability to provide feedback on the results – for example, whether a prediction correctly identified an anomaly or not – and Lookout for Vision will automatically retrain the underlying model.
Lookout for Vision is available directly via the AWS console as well as through supporting partners to help customers embed computer vision into existing operating systems within their facilities.
Industrial vision provider, Basler, is offering its customers complete vision solutions using Lookout for Vision. The firm's head of marketing, Gerrit Fischer, said: 'Basler and Amazon Lookout for Vision provide a very lean architecture to adopt vision-based anomaly detection in any manufacturing site. We’re excited to jointly provide our customers with complete vision solutions by combining Basler’s expertise in industrial vision and edge platforms with AWS’s investments in industrial machine learning.'
Other Lookout for Vision users include GE Healthcare and food manufacturer Dafgards in Sweden. Fredrik Dafgård, head of operational excellence and industrial IoT for Dafgards, commented: 'We previously tried Amazon Lookout for Vision to automate the inspection of our pizza production lines to detect whether pizzas had enough cheese and the correct toppings, with good results. We’re excited to extend Lookout for Vision to our other production lines such as hamburgers and quiches, to help us detect any anomalies like incorrect ingredients. Over time, we plan to scale Lookout for Vision across multiple production lines.'
Kozaburo Fujimoto, operating officer, general manager, manufacturing division, plant manager at GE Healthcare Japan, commented: 'As one of the world’s most trusted healthcare companies with more than a century of technological progress and digital innovations, we look forward to capitalising on the benefits that AWS’s industrial machine learning services will potentially bring to our manufacturing environments.'
