Visual inspection with AI: how do you implement it?
Visual inspection has a lot to offer the maintenance world. Condition measurement of assets remotely saves time and money and is good for safety. We talked earlier about IBM Maximo Visual Inspection. In this blog, we provide some examples of practical applications of this software for visual inspection with AI. What can you do with it? And what can Gemba do to help?
Offshore: visual inspection of unmanned platforms
Suppose you need to check the condition of the legs of an unmanned gas or oil platform. A boat or helicopter with maintenance personnel alone is not enough. For example, a medic is also needed and someone for communication with the mainland. A costly operation that also requires planning. It is also not without danger to have people ‘abseil’ along the legs on a rope. Instead, you can also send a drone that takes pictures of the legs and then analyzes these images with IBM Maximo Visual Inspection. How rusty are the assets? And where is maintenance needed? By feeding the system with more and more images and indicating where the pain points are, you teach it to recognize them independently.
Infrastructure: visual inspection of tunnel lighting
The maintenance of our traffic infrastructure often takes place on the basis of performance contracts. The longer a road, tunnel or bridge is closed for maintenance, the more money it costs the maintenance party. With IBM Maximo Visual Inspection it is possible to limit this. For example, consider a moving car with a camera on the roof that takes pictures of the fluorescent lights in a tunnel to determine whether they still meet the standards. By having someone with knowledge of the matter label the images in IBM Maximo Visual Inspection and indicate what is acceptable and what is not, the system is trained to determine when a lighting unit needs to be cleaned. This way, the tunnel can remain open during an inspection of the lighting and that yields a lot of profit.
Network management: visual inspection of hazardous installations
In addition to images from ordinary cameras, you can also analyze X-ray and thermal images with IBM Maximo Visual Inspection, for example. An example of a possible application for grid operators is the continuous monitoring of components in high-voltage or medium-voltage stations with an infrared camera. Here too, you teach the system to determine where maintenance or replacement is needed. This way you don’t have to put employees to work in a dangerous environment.
Production, construction, gas and oil, industry: visual inspection for safety
With IBM Maximo Visual Inspection it is possible to increase the safety of the working environment in various ways, and this applies in fact to all sectors in which the risks are greater than average. With a webcam, for example, you can continuously take pictures to check whether your employees are wearing a helmet or other required personal protective equipment. Or think of automatic alarming in a ‘man overboard’ situation or if an employee suddenly hangs still during maintenance on a storage tank. You train the system to recognize these types of dangerous situations. The new European Corporate Sustainability Reporting Directive (CSDR) makes monitoring and reporting on employee safety more important than ever.
No data scientist needed with IBM Maximo Visual Inspection
For whatever application you use IBM Maximo Visual Inspection, the big advantage is that you don’t need a data scientist to train the system. The software writes the data model and logic itself. A subject matter expert can independently assess the images and indicate what the points of attention are. By having this done not by one person, but by a team of substantive experts, you bring even more objectivity to the findings.
Step towards predictive maintenance and asset performance management
Visual inspection is currently being used by companies on a limited scale. For example, they take pictures with drones and write a report about it. But these images are not used to feed a system such as IBM Maximo Visual Inspection with data and thus determine the health of an asset or the ‘asset health’. While this is precisely so important to optimize maintenance and take the step towards predictive maintenance. Asset Performance Management (APM), as we call this process of monitoring assets, determining asset health and predicting maintenance needs. It helps you make smart decisions about the maintenance of your assets and save time and money.
Demo app IBM Visual Inspection
Although AI-driven maintenance applications may still be used to a limited extent in practice, they are undoubtedly the future. At Gemba, we help you implement IBM Maximo Visual Inspection in the right way and link it to IBM Maximo Manage or other maintenance systems. We are busy with it both behind and in front of the scenes. This ranges from obtaining the necessary IBM certifications and creating proof of concepts for a practical application to organizing workshops on the possibilities of IBM Maximo Visual Inspection for employees and customers. And we at Gemba all have a demo app for visual inspection on our mobile phones. With this app it is possible to inspect the condition of miniature solar panels by analyzing images in real time. At the Maintenance Next trade fair, last April, visitors enthusiastically experimented with this demo app.
Would you like to discuss your asset management challenges?
Learn more
Would you like to know more about IBM Maximo Visual Inspection or see with your own eyes how the demo app for visual inspection works? Please contact Gerben ter Horst via +31 (0)6 - 14 57 21 19 or g.terhorst@gemba.nl.
