Our takeaways from the NLAIC conference
Yesterday we had the pleasure of attending the NLAIC conference in the Jaarbeurs in Utrecht. The entire event was centered around getting to an implementation stage in AI development. A lot of energy and time has been poured into a wide variety of AI projects over the past couple of years. It’s really time for these projects to move beyond a proof of concept (PoC) phase. All the hype around AI is lovely of course, but we will only be able to reap its benefits when we actually implement it in a production environment.
Productionizing AI is hard
At UbiOps we’ve been focussed on getting AI/ML projects from a research stage into a proper production environment for years now, and lots of the struggles shared yesterday were no surprise to us. Going beyond that PoC stage comes with a whole set of hurdles to overcome! From data access and availability, to scalability issues. It was nice to see that more people are aware of these issues and that the NLAIC is making sure people join forces to overcome them together.
Predictive maintenance and digital twins have a lot of impact potential
Some projects also showed successful implementations already. We were especially impressed by the digital twin set-up demonstrated by FreezerData and the HvA, which is used to help the development of predictive maintenance tools for hospital freezers. With their digital twin they managed to generate enough data to create early maintenance warnings. This significantly decreased the down-time of these hospital grade freezers, making sure that critical resources like blood and vaccines were kept safe.
There appears to be a lot of impact potential for predictive maintenance in the technical industry. The workgroup “technische industrie”, led by Patrick Blommerde, recognized this and is spearheading several projects in this area. Currently the technical industry is dealing with a shortage of maintenance technicians and very high costs for downtime of any instruments. Predictive maintenance would enable a small group of maintenance technicians to do their work efficiently and effectively, decreasing down time and associated costs.
At UbiOps we’ve been working on productionizing predictive maintenance cases and digital twins for several years now. If you want to read more about this have a look at our AssetRail and Gradyent use cases!
Collaboration is key in AI development
Without collaboration between different companies and institutions, it’s very difficult to take AI to the next level in the Netherlands. With everyone working in silos we are doing a lot of double work and our innovation speed slows down. The NLAIC is actively bringing organizations together from across the Netherlands, to join forces. Several “chain projects” have already been started through the NLAIC to make sure that AI advancements benefit a whole value chain, rather than just a single company. The NLAIC also has a joint approach with the AiNed programme to help get AI projects funded. It’s wise to keep an eye on the AiNed website for any opportunities!
Do you have any AI projects you want to undertake and do you still require a technical partner to take care of the ML operations side of things? Make sure to reach out to us to see if we can be of help.
A great conference with a clear call to action
Overall the NLAIC conference was a great event. Many interesting speakers like Eppo van Nispen tot Sevenaer, and Kees van der Klauw. All had one clear message “Let’s embrace the AI opportunities in the Netherlands and help each other to actually implement the right tools, in an inclusive fashion”. It’s a message that we wholeheartedly agree with. Our mission is to help teams build and manage scalable AI solutions, to drive better business outcomes and a better world. AI projects have no value when they are archived away after a proof of concept! AI is not something that should only be accessible to the big tech companies. Let’s join forces to properly productionize AI together, shall we?
If you have any questions about what else is possible with UbiOps, or you’re interested in deploying your model to UbiOps, do not hesitate to reach us and book a call with our Product Specialist!
