NEWS
Amazon reveals how it uses AI to spot damaged products
M
illions of products pass through imaging tunnels in Amazon
fulfillment centres each and every day. Amazon’s product categories now range from large electronics and equipment, to fresh consumables delivered in under an hour. Organising its products into an efficient and reliable distribution system has always been at the heart of its success, but with so many products of varying size, shape, weight and packaging, artificial intelligence (AI), combined with computer vision, are crucial. Project P.I. uses detective-like tools to scan items and uncover defects in products or packaging, such as the incorrect colour or size, before a product reaches the customer, helping to reduce returns. When an inconsistency is found, the tool then helps to identify the root cause of issues, enabling preventative measures upstream to stop it recurring.
Amazon’s AI detective, Project P.I
All Amazon products are passed through an imaging tunnel before leaving the fulfillment centre, and it’s here where Project P.I. uses computer vision to scan the product, then evaluate the images for any defects. If a discrepancy is found, the AI investigates further, determining how deep the issue goes. “We want to get the experience right for customers every time they shop in our store,” said Dharmesh Mehta, vice president of Worldwide Selling Partner Services at Amazon. “By leveraging AI and product imaging within our operations facilities, we are able to efficiently detect potentially damaged products and address more of those issues before they ever reach a customer, which is a win for the customer, our selling partners, and the environment.”
organiser of the International Machine Vision Standards booth, which is located this year in Hall 8/Booth No. 8E20.
By Thomas Lübkemeier, EMVA General Manager
EMVA Activities around VISION 2024 EMVA is pleased to organise the popular International VISION Night event on the evening before the VISION 2024 starts in Stuttgart, Germany. During this networking event, the machine-vision community will meet for dinner. We look forward to welcoming exhibitors as well as visitors of the 2024 VISION show. More details on the event and how to register for the International VISION Night can be found at
www.vision-night-
emva.org. Premium sponsors of the event are Messe Stuttgart and the US-based company Birger Engineering. During the three trade show days of VISION 2024, the EMVA once again is the leading
Seventh European Machine Vision Forum – Where Research Meets Industry
‘Challenges and Chances in Computer Vision for Human- Machine Interaction’ is the focal topic of the seventh European Machine Vision Forum, from November 7 to 8, at the École Nationale Supérieure d’Ingénieurs Sud- Alsacein (ENSISA) in Mulhouse, France. Machine vision is a promising approach to enable the interaction between humans and machines. The forum is organised by the EMVA and aimed at scientists, development engineers, software and hardware engineers, and programmers both from research and industry. More information and registration at
www.european-
forum-emva.org.
6 IMAGING AND MACHINE VISION EUROPE DECEMBER 2022/JANUARY 2023
Environmental impact of defect-detecting AI
All this effort to check and re- check product quality isn’t solely for the good of the customer. It’s also a large tool in helping Amazon to deliver on its sustainability goals for the future, too. While preventing damaged or defective items from reaching customers has clear benefits to the customer experience, it
also results in fewer returns of unwanted items, meaning less wasted packaging and overall carbon emissions. “Amazon is using AI to reach our sustainability commitments with the urgency that climate change demands, while also improving the customer experience,” commented Kara Hurst, vice president of Worldwide Sustainability at Amazon.
Amazon’s Project P.I. system uses detective-like tools such as computer vision and generative AI to scan items for defects before sending them to the customer, helping to improve its systems and reduce returns
New EMVA members Since spring, four companies have chosen to become an EMVA member. These include N.A.T. – a company that is creating sophisticated state- of-the-art solutions for wired and radio based applications in telecommunication and industrial automation. Founded in 1990 and based in Bonn, Germany, N.A.T. is one of today’s leading-edge suppliers for board and system level products, as well as turnkey and application- ready systems. Our second new member is Inno-spec, which is based in Nuremberg, Germany. The company develops and produces spectroscopic measuring devices for industrial use. The focus is on hyperspectral cameras for recording spatially resolved spectral data. Together with its partners, Inno-spec provides its customers solutions from components to integrated solutions. The third new EMVA member is Delta Electronics, a
global provider of power and thermal management solutions headquartered in Taiwan. Part of its product portfolio is a machine vision system series that was developed to meet the demands for industrial automation applications on the production line, such as stain inspection for product quality check, exterior size measurement, finished products count, identification check and other industrial automation related processes. Lastly, Shenzhen Do3think Technology has joined the EMVA; it is an industrial camera manufacturer established in 2007 and specialised in industrial camera R&D. Do3think line scan and area scan industrial camera series are widely used in consumer electronics, PCB, semiconductors, automobiles, logistics, photovoltaics, lithium batteries, medical care, life sciences, food, packaging, printing, textiles, agriculture and other industries.
Amazon
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39
