Simplifying visual inspection

IBM Maximo application and iPhone deliver defect detection at a fraction of the cost.

Visual inspection is an opportunity. It is estimated that the market will be worth more than $25 billion worldwide by 2027. Companies are excited about the prospect of cost-savings, error reduction, better accuracy, and higher quality output. Automated quality assurance is the utopia all manufacturers want to achieve. The pandemic helped the process along, as companies tried to reduce the amount of human intervention. But finding the right skills and experience to manage a visual-inspection system can be challenging and it is an expensive investment.

Visual inspection is an application of computer vision with machine learning and artificial intelligence that helps humans make better products with fewer defects. It is much more efficient to find the defects during the process rather than fixing issues later or having to initiate recalls.

We have been working with Professor John Ward and his team at the University of South Carolina on a new way to perform visual inspections. Rather than waiting months for a system to be designed and installed, Professor Ward has created an easier answer. This leverages the IBM Maximo iPhone application as an input device to create a visual model using photos. The iOS devices are installed on the production line.

There is a beauty in the simplicity of using smartphones, but any visual-inspection system creates significant amounts of data. High-speed, 5G, millimeter wave, low-latency networks have to move high-resolution images to a system to analyze the information and communicate results in split seconds to stop a line in case of a defect. Having a system to collect information is only the first step. The power is in the data.

Our contribution is a comprehensive business-intelligence analytics suite that allows companies to drill down all the way to see the image of the defect. It means companies can look at the return on investment (ROI) per defect caught as well as all the way up to the line or plant or number of plants so the company can start aggregating quality data and use it more broadly. Instead of being a point solution, the visual inspection is now a company-wide node in a manufacturing machine.

We also bring an agile way of working and a focus on collaboration, iterative development, and continuous learning. This enables the rapid delivery of business value so ideas get to market faster. In addition, we further amplify the impact of agile methods by providing specialized expertise in the latest technologies, including cloud, automation, artificial intelligence, and high-speed private networks. From idea to launch, the advanced technology delivery team working from the development center in Columbia, South Carolina, helps visionary companies turn ideas into business value at industry-leading speed.

Dashboards make managing corrective actions easy. This can include identifying areas that need more training, since people are a huge variable in building products. The immediate feedback of real-time data makes operators more efficient. It also paves the way to a continuous improvement culture and mindset.

Visual inspection is about more than just quality; it can also be applied to sustainability initiatives that seek to decrease waste and increase accountability through managing more variables at greater levels of detail. As manufacturers trend towards Industry 4.0 with smart plant-control towers, private 5G, and edge computing, this visual-inspection application can provide the rationalization for these foundational technologies to finally present a clear business case for adoption.

Find out more about our partnership with IBM.