Key ingredients for successful AI implementation
Implementing an AI strategy across your supply chain can be overwhelming, but Matthias provides us three key ingredients for a successful implementation. You need to have:
- A strategy for AI compute architecture — this involves deciding whether you’re going to build AI infrastructure in-house, or rely on third party cloud services.
- A robust data strategy — this is important to provide a solid foundation for training AI models
- The right people— you need talent that can translate your actual business problems into the requirements for an AI model.
According to Matthias this third element is the most critical. “You need someone who understands your business well enough to be able to say, okay, this is the problem that we're trying to tackle first when we implement AI. And then obviously you need talent that can take that translation and turn it into an actual model implementation” he says.
The current state of AI in the supply chain
We are not yet at a point where we can say AI can solve everything in a better, faster, cheaper way. Matthias explains that we are currently pushing the boundary of research to ensure that the domain experts who will end up using these models and methods will trust them. “Developing AI methods or models that would be able to beat existing state-of-the-art methods in terms of solution speed, solution quality, but most importantly solution cost” he says is where we are headed.
We’re still at the beginning of AI’s adoption across the entire supply chain industry. Currently its cutting-edge technology requires resources that not many companies in the industry have. But Matthias envisions a future where this technology becomes more widespread across the industry as it becomes more accessible.
What to watch out for
It’s easy to get caught up in the excitement of AI and the benefits it will bring, but it’s power is not fully understood yet, so Matthias advises caution in three main areas.
- Be careful with confidential information
- Be mindful when investing into the resources and talent needed to get AI models off the ground
- Be aware that building these capabilities takes time.
Striking a balance
“Yes, AI and machine learning is probably the next big thing in the supply chain logistics industry, but it's not the only thing you should be thinking about” says Matthias. He encourages companies to find the right balance between investing some of their budget for research development analytics into exploring this space, but not all of it.
AI has an important role to play in the future of the supply chain industry, and Matthias believes there is still a lot of research to be done to optimise the technology for organisations.
