On its homepage, the Hannover Messe already announces the trends for 2022:

“Networked robots are entering the smart factory, new materials and processes are turning traditional manufacturing methods upside down, and calls for more efficient processes are growing louder. Integrated industry is permanently changing.” From our daily interaction with our customers, we can fully underwrite this. We have compiled a list of AI-based industry implementations that are trending and what the ultimate Industry 4.0 toolset could look like.

1. AI and the greater cloud

2. MLOps

3. AI-driven visual inspection.

4. Anomaly detection increases productivity

As AI is taking over industrial applications, it has a significant effect on the adoption of Cloud Solutions. The deployment of AI makes it possible to monitor and manage cloud resources and its vast amount of available data. Artificial Intelligence is said to bring increased value to the cloud, which is often seen as no more than an optimal solution for data storage today. An AI cloud can bring together the AI hardware (edge devices) and software to offer Managed AI services(SaaS-like AI solutions). This is also the journey that we have seen for navio, our machine learning operations platform that is available on Azure as well as on-premise. Meanwhile, we have partnered up with TTTech Industrial to offer users the ability to deploy their fully managed machine learning models directly to edge devices. Integrated with the Nerve Management System, users can now directly execute their models on their connected Nerve devices. Our customers can get predictions in real-time directly at the edge, or off-site, from their models running in the cloud.

This results in:

• reduced latency

• improved security and

• better performance

When having a look at the search volumes of MLOps, you will see that the keyword had almost no traffic two years ago; since then, the search volumes have exploded! Over the years, a company’s IT system has increased in complexity. As such, customers are requesting more and more platform solutions that combine every step of the machine learning process: deployment, training, and monitoring. Also, the infrastructure needs to be relieved. IT operations and other teams can improve their key processes, decision-making, and tasks with MLOps solutions and improved analysis of the volumes of data coming its way. Many business leaders are on the lookout for an end-to-end solution, like craftworks’ navio, that empowers cross-team collaboration, and brings ML models out of the lab and into production.

New research reveals that, what is driving the adoption of AI in manufacturing, is the pending need for business continuity in the face of a changing business environment. The pandemic led business owners to turn to disruptive technologies and, in a way, spurred innovation in manufacturing. 66% of manufacturers who already use AI in their day-to-day operations report that their reliance on AI is increasing, says a Google Cloud report. Another common use case is overall employee efficiency with the help of prescriptive analytics in the form of real-time guidance, flagging safety hazards, or detecting errors on the assembly line. Moreover, the sub-sectors of these adopters are automotive/OEMs, automotive suppliers, and heavy machinery. This report also underlines what we have observed so far with our customers. That is the need for visual inspection of a finished product.

For product and service improvement, manufacturing leaders have kept an eye on predictive maintenance and predictive quality. Innovative solutions in the manufacturing industry arrived in the form of a combination of edge devices + anomaly detection algorithms that can predict maintenance needs. Comparing an AI-based approach to traditional condition monitoring or more classical maintenance, a considerable improvement is expected due to better failure prediction. In some cases, availability can increase by more than 20%, while inspection costs may decrease by up to 25%, with an overall reduction of up to 10% of annual maintenance cost possible. In our experience, implementing a pilot for an AI-based anomaly detection showing concrete improvements can already be achieved in 6 weeks. Also, the production implementation and rollout is a matter of months and not years.

These are only four prominent topics while developing AI-based industrial process optimization solutions. There are many more. Join us at the Hannover Messe and let us know what other needs you have for AI-based applications in your industry. Don't have a ticket yet? Get your free Hannover Messe ticket here!

You can find us here: Hall 3, Stand A20, (B25)