The COVID pandemic is changing many aspects of work. Remote access has gained acceptance in many industries and is changing the nature of work from the plant floor to engineering departments. In last year’s workshop on the challenges of AI adoption, it became obvious that technology is only one of many challenges. Much more critical for the success of AI introduction are human factors and organization.
At this year’s Forum, the AI workshop will look at the interdependency between humans and AI. We will focus on industrial applications including:
- Engineering: Where is AI used in engineering to keep projects in time?
- Service & maintenance: How can AI support the maintenance staff remote and on-site to be more efficient?
- Troubleshooting: How can AI support the operations team with trouble shooting, when a machine stands still?
Most of us use AI every day. It is the auto correct / auto complete phrase on our smartphones. In ARC’s view, there is a great potential for AI in the industrial world in exactly these use cases, such as auto complete for PLC programming with instant coding assistance, automated testing, etc. Often machine builders need to adjust their programs slightly to fit a customer specific machine requirement. AI can help here to adjust the code quickly. A special case may be the translation from a Rockwell to a Siemens PLC code, something that still can’t be done.
Service & maintenance
Besides trouble shooting, see below, among the most obvious examples for AI is to use it in predictive maintenance. AI can also support the maintenance team with suggested measures or simply with using route optimization between tasks (travelling salesman problem)
Think of the following scenario. In the middle of the night a crucial machine suddenly stands still. The operations team is flooded with alarms. The resources are limited. There is little chance to call your local service partner or machine builder. Each minute of downtime is costly.
There are various ways AI can help here. The first is obviously alarm management, where AI can help to trace the source of the problem. If you have a hard time imagining this, watch the movie passengers, AI could have helped Jennifer Lawrence to track down the original problem in seconds. Some further reading: https://dzone.com/articles/how-aiops-revolutionizes-alarm-management
AI is also already used in problem solving. Years ago, I visited an IBM booth at the Hanover show, where they showed who Watson is used to support trouble shooting. On our last forum, it was also discussed and Pfizer showed us their solution for maintenance teams.
All these examples have one thing in common, they need a tight “collaboration” between AI and human workforce. AI can do a lot, but it cannot hold a screwdriver or fix a problem, where we do not have data. The closer integration for AI also bears challenges for the worker/operator/engineer. How much AI should be integrated? How much is accepted by the workforce without feeling infantilized by a machine?
In the workshop, we plan to exchange practical examples and discuss the future development.
Picture Source: Pixabay