Jun 15, 2022
SmartMore participated in Manufacturing Expo held in Birmingham, UK, during 8-9 June 2022, where its smart industrial platform (SMore ViMo2.0), barcode scanner series (SMore VS1000), and quality control solution were presented.
Manufacturing Expo is an exhibition for global manufacturing professionals to maintain a competitive edge by discovering the latest innovations, technologies, and solutions. SmartMore not only presented the smart products, such as SMore Industrial Platfrom and SMore Code Scanner VS1000, but also attended a workshop to demonstrate the smart solution of quality control.
How SmartMore’s solution solves the last-mile problem?
- Deep learning, instead of the traditional machine vision
When it comes to counting items, sorting items out, detecting defects, etc., the traditional machine vision frequently applies. But this traditional approach is beneficial only in a straightforward manner.
The limitations of traditional machine vision are: What if the defects come in more than one type? Or what if the production line change from one type of product in the morning, to another type of product in the afternoon. That means manufacturers need to apply new models to any changes, costing more money and time.
On the contrary, deep learning models are more flexible, because such models, just like human brains, can learn patterns and features on their own. This enables manufacturers to apply deep learning models from one product to another, and from one type of defect to many other types of defects.
- Algorithms specialized for manufacturing industry
SmartMore’s quality control solution is built on its smart industrial platform (SMore ViMo) with four major algorithms: detection, classification, segmentation, and OCR.
Such algorithms can help manufacturers detect a multitude of defects precisely, classify between OK/NG products, capture the region of interest for better inspection, and recognize optical characters under harsh conditions.
- Flexible deployment
The traditional machine vision approach usually features a long cycle. The operator usually needs two days to collect, analyze, and label data, and then algorithm engineers also need two days to train models. Next, software engineers will release the models in the format of SDK, which may take 7-14 days.
SmartMore’s solution indicates a shorter cycle instead. After the operator collects data, the whole process of data labeling, model training, and model release can be completed within one single day.
Traditional machine vision approach VS SmartMore’s solution:
SmartMore will also attend another two exhibitions of manufacturing held in Europe and Malaysia, which are automatica Munich 2022 (21-24 June) and METALTECH & AUTOMEX 2022 (22-25 June).
Get your free ticket to meet our team on the ground and learn more about our products and solutions:
- automatica Munich 2022: Get Your Free Ticket for automatica Munich 2022 (google.com)
- METALTECH & AUTOMEX 2022: METALTECH & AUTOMEX 2022 Registration (eps.net.my)