Inspekto has launched a new software version of its INSPEKTO S70 autonomous machine vision system. Based on customer feedback, the S70 now offers smart features that guide users in creating and maintaining inspection profiles, and improves usability, process integration, and accuracy of inspection.
INSPEKTO S70 mimics cognitive human vision with three synergetic AI engines driving the unique electro-optics:
• The first AI engine provides acquisition and optimises hardware settings based on the inspection scenario and dynamic environment.
• The second AI engine provides detection to recognise and analyse the part under inspection.
• The third AI engine then inspects the part, looking for defects that may be present in the object.
The inspection of the part is completely autonomous. The system can detect defects, with no prior defect definition required; the AI engine can recognise the characteristics of a ‘good’ part and based on this information, can deduce whether another part is defective. This AI provides many benefits, including:
• Removing the time-consuming process of creating defect definition rules.
• Being able to detect small defects that would be missed by the human eye.
• Being able to detect defects on a black surface.
• Allowing for inspection of complex 3D shapes.
The recent upgrade to the INSPEKTO S70 system now incorporates a new Profile Center to help users improve and optimise profiles over time. This is a set of smart tools that guide users when adjusting a profile, to reach the desired performance level and ensure that it is adapting to changes in the production process and environment such as tooling change, component placement or lighting variations.
The Profile Center also allows for the comparison between previous profiles and new profiles for the same item to be able to continuously improve the inspection performance. To ensure continuous inspection performance throughout the life cycle of the inspected product, the S70 is now able to autonomously generate AI-based active recommendations which adapt the profile to production changes, either process- or environment-related.
Users have the option to increase or decrease the sensitivity of the system to specific types of defects, while sensitivity to other defects, even unforeseen ones, remains unaltered. This can be done by selectively adding defected (NOK) samples to the profile parameters. The new version of the INSPEKTO S70 allows users to add to the profile, at any given time, both good (OK) and defected (NOK) samples for performance finetuning.
This reduces false alarms for NOK products, and faulty products incorrectly flagged as being OK. The adding of samples to the profile is immediate and results in a profile that has been optimised for the specific part being produced.
Regions of interest (ROIs) within the part can now also be defined, and the size of the region and the sensitivity thresholds can be adjusted independently for each part.
“Currently, every time a company chooses a new supplier for a component, this disrupts the inspection set up and a machine vision expert must reprogram the software. The new version of our software eliminates these challenges, allowing the plant personnel to ensure the profile’s continuous performance to achieve top notch inspection, without relying on external experts,” explained Bernd Schumacher, global CEO of Inspekto.
| Tel: | +27 11 609 1244 |
| Email: | [email protected] |
| www: | www.zetech.co.za |
| Articles: | More information and articles about ZETECH ONE |
© Technews Publishing (Pty) Ltd | All Rights Reserved
printer friendly version