Jidoka Technologies, a firm in the field of automated cognitive inspection for manufacturing, has recently launched Kompass 3.0, a next-gen AI-enabled visual inspection solution. The updated platform is a result of Jidoka further strengthening its technological innovation capabilities. This latest version of the solution has made the organization rapidly respond to changes in the market and customer requirements, thereby giving customers the much-needed competitive edge.
Dr. Krishna Iyengar CTO and Co-founder, of Jidoka Technologies, said, “Kompass 3.0’s core product has made significant advancements in three areas, realtime insights into quality, new solution accelerators, and time to go live, speed and robustness. The robustness of inspection has increased to 4,200 parts per minute from the earlier 500 parts per minute. For very complex systems where there is a need to augment human decision-making, the human-in-the-loop AI approach is taken that brings together human intelligence and AI for the creation of machine learning models.”
Key features of Kompass 3.0 include,
- Advanced Dashboard to give real-time insights into quality in the production, logistics, and print workflows.
- Vertical Solutions such as Print Inspection Accelerator and Logistics Applications Accelerator have been added to the core product. This enables warehouses to ensure accuracy and quality in inbound and outbound processes. Similarly, in print inspections, validation of dynamically changing print data is done at speeds of up to 3600 items per minute.
- Anomaly-based decision systems provide an opportunity for companies that are establishing new products or lines where very little is known about defects to leverage the AI capability of inspection. This is in addition to the object detection and classification models that are available.
- In addition, it gives customers the ability to write custom code modules to enhance the effectiveness of the solution while providing flexibility without losing robustness.
Sekar Udayamurthy, CEO and Co-founder, Jidoka Technologies, said, “Since our first launch of the solution in 2018, we are continuously upgrading the Kompass Platform. The latest version of the platform provides new as well as existing customers with a more powerful visual inspection system that caters to the changes in business and quality requirements of the manufacturing industry.” He further added, “This latest version helps with traceability, real-time insights, and human-in-the-loop decision-making, giving customers control to balance quality and throughput and other benefits such as robustness, speed, and being more hardware agnostic. This development was possible because of our team’s constant endeavour to develop future-ready solutions.”
Jidoka Technologies’ patented software platform, Kompass is an AI-powered defect detection and analytical software. It enables manufacturing companies to detect defects visually in real-time. It provides a turnkey solution that seamlessly fits into the production line delivering up to 4200 inferences per minute. Several leading automotive, FMCG, and Pharma organizations are already leveraging the power of the solution and enhancing their output efficiency.
Customers in the manufacturing industry can now deploy and manage the solution without a higher level of technical expertise and skills. The software can integrate with MES and ERP systems and provide an operational report for day-to-day management. It can increase throughput, speed of inspection, and defect accuracy. The solution reduces excess wastage, provides insights into process threads, and brings down fatigue and stress levels of the Quality Control Teams. It can help the Subject Matter Experts in upskilling themselves by taking their focus away from mundane quality checks to a continuous recalibration of defects. Visual Inspection resources and Human Resources Costs can be reduced significantly too. The manufacturers realize their investments in less than 12 to 24 months while upgrading to a more streamlined quality check process.
Jidoka’s R & D team is also working on developing next-generation technology that includes Self-Learning. For products with high variety and frequent design changes, this technology will be in demand. This next-gen AI Tech will deliver anomaly detection for unknown and unseen defects. Here AI “understands” what normal looks like and “imagines” it where the input image is “healed” to represent the “OK” component.
Jidoka solutions deliver greater than 98 percent accuracy in defect detection in the QC process with a significant increase in throughput by 33% and a decrease in false positives by as much as 30-40%. A 20% increase in the speed of inspection, with 5-8% further accurate detections, is made possible with the solution.