Starting steps for collaboration with AI

Introduction

Manufacturing is a complex domain. Bringing Artificial Intelligence into manufacturing is no simple task either. When Wärtsilä’s AI Center of Excellence was called to join the DIM venture together with other colleagues from Wärtsilä, little did we know where we would be after only a short period of time. As of now, we are looking at multiple AI applications explored, piloted and ready for handover for industrialization, some that we were not even thinking of. The main recipe has been cross-functional collaboration within Wärtsilä’s Manufacturing Acceleration Centre, ecosystem partners, and just doing stuff.

Sometimes just by doing, showcasing, and communicating, a vision becomes tangible. Additionally, Wärtsilä has been adopting a new method of learning by doing and collaborating, Datamob.

From doing stuff to vision

Our initial idea to start exploring AI solutions for manufacturing was to apply artificial intelligence to assist assembly workers. We started working on this problem and quite soon got the gist of some possible solutions. One specific example was to highlight the workflow of the given assembly activity in corresponding engineering drawings. We moved quickly: Found data, labeled it, trained a model, and gathered feedback from assembly workers. The solution is now being integrated into the process.

Yet, that was not it. A simple use case took an interesting turn and started to evolve. In another cross-functional collaboration, this solution was presented and an idea to apply this method for quality assurance of drawings was proposed. This occurrence opened the door for multiple new ideas to apply artificial intelligence to our design documentation. From one solution built from the ground-up, we are currently looking at a growing toolbox of AI solutions for design documentation to be applied throughout the design lifecycle. We had no grand vision for AI, no executive PowerPoint, just people doing things and learning as they go, working toward a future where every person and system has the insight they need to act with confidence. Along the way, AI has proven to be a practical accelerator, helping us move faster from ideas to impact. Datamob: breaking through the pain barrier, together.

Collaboration and learning by doing can be difficult to start and the key is to take the breaks off and just try things together. In Wärtsilä’s Manufacturing Acceleration center we have piloted a new approach in rapid innovation and collaboration: Datamob. In a Datamob, we create teams of colleagues with different expertise to work on real-life data problems. Usually a team of five, the team works together on one laptop, constantly changing roles. Every five minutes, a different person drives the problem solving in the laptop together with a navigator team member (and quite often an AI tool in a vibe coding manner). Meanwhile, the “back bench” can brainstorm and think of the overall direction of the problem solving. This forces people to adapt quickly and break through the pain barrier. It does not matter if you do not know how to code, you’ll figure out something in five minutes when it’s your turn! Also, struggling together creates a fertile ground for good team spirit. Coming up with a perfect solution is not the end goal. It is collaboration, learning by doing and rapid innovation. Getting teams to grasp new problems together can lead to new ideas in other avenues. Also, quite often, finding sub-problems in a problem can be even more valuable than solving the original problem.

This spirit of experimentation and teamwork is what we aim to carry forward.

Looking ahead

As we continue exploring AI in manufacturing, one thing is clear: innovation doesn’t come from perfect plans – it comes from people working together, learning fast, and trying things out.

Next, we will host the first Datamob with partners on September 23rd. In the future, Wärtsilä is happy to invite others to join datamobs and other practical collaboration.

 

Tony Savander
Senior Data Scientist

Wärtsilä

 

 

Andrea Kovalova
Junior Data Scientist

Wärtsilä

 

 

Maria Litova
Junior Data Scientist

Wärtsilä