We ask these 2 experts how the Internet of Things, Artificial Intelligence and Machine Learning are changing the Supply Chain profession. Buckle up for a conversation at breakneck speed.
We have been hearing and reading for a long time about how the Internet of Things and other recent technological developments are causing a landslide in the Supply Chain planning profession. What is this all about?
Joël : (smiles) « Good afternoon. Indeed, there is no shortage of enthusiastic reporting. But honestly: it is really impressive to see which practical applications can be devised with the new technology in the broad field of Supply Chain and operations Really impressive. »
But how does that manifest itself in concrete terms?
Joël : « Well, our profession, Supply Chain management, is all about creating transparency throughout a logistics network. Today more than ever, you have a strategic advantage if you have a clear view of the increasingly complex globalized Supply Chain networks. »
Bart : (nods) « Due to faster access to information, we can respond more adequately to a number of typical logistical challenges: late deliveries, last-minute orders, goods that are stuck in customs and so on. Data makes that reality visible more quickly. »
Joël : « But it goes further than that: new data sources allow us to structurally better organize the chain: sensors that transmit information during transport, updates of stock positions of your most important suppliers, live information feeds about customer sales or about the progress of production orders. But also think of updates from carriers. Via the cloud and IoT (Internet of Things, ed.) A mass of information is generated in real time, integrated and made clear. The dream of every Supply Chain manager. »
So creating a better view of the end-to-end chain, is that the most common application you see?
Bart : « Well, visibility is an important first step, indeed, but with the new technology we can also automate a lot of planning and coordination work, or at least simplify it... Today we see that in many organizations, people are really putting a lot of time and energy into manually creating and updating sales forecasts and production plans. Spreadsheets are flying around, across the organisation and employees lose themselves in manual data work and constantly re-solving the planning puzzle. This requires a tremendous amount of energy and many organizations really want to get away from it, but they do not always know how. There is still a lack of insight into how this can be done differently. »
And how exactly can technology help with this?
Bart : « For example in forecasting, i.e. predicting future customer demand as well as the resulting need for products and resources. We develop algorithms that combine historical sales information with data that we collect from the Internet, such as weather information or traffic information. Or other sets of data that we think influence how demand will evolve, in the very short or long term. An intelligent algorithm automatically recalculates the forecast at an aggregated or at a more detailed level. The clever thing about this is that the algorithm has a self-correcting feature: it learns from its mistakes and it looks for new causal relationships. Over time, it makes increasingly better predictions. So Machine Learning. »
Joël : « But inventory management and production planning are also the perfect domain for such software. Artificial intelligence to support the planner. The perfect combination as far as we are concerned. »
Does man still have a place? Or will we soon be replaced by intelligent algorithms in the cloud?
Bart : « Technology facilitates decisions, but people remain in control. For example, our prediction model make a proposal for the planner to review… The planner can agree or not. Today we see that planning professionals agree in 70 to 80% with what the models propose (smiles). As humans, we still have to get used to this, because we like to keep control. But what we do notice with our customers is that people often quickly see that the algorithms make proposals which are sharp, substantiated and focused. Of course, change management and transparency are very important in accepting and fully embracing these solutions. »
Joël : « But as confidence in the new technology grows, more complex challenges are being considered. Challenges such as setting up intelligent planning solvers that enable advanced inventory optimization in multi-echelon networks. Due to these new possibilities, the role of the planner can gradually shift towards more tactical work. Evaluating and adjusting the plan, no longer the daily cutting and pasting work in spreadsheets or manually aligning ERP data with planning systems and vice versa. Please note, there is still a lot of knowledge that is not contained in figures or data. Not all insights and know-how can be derived from data sets. And that is precisely the added value of planning professionals, as far as we are concerned. The technology should relieve people of the data hurdles that make life difficult for them today, so that they can concentrate on reviewing what the system calculates. That human touch is worth its weight in gold. »
But wasn't that also the promise of ERP and planning software over the past two decades?
Joël : (laughs) « I know .. This message has indeed been conveyed in the past, when ERP and planning packages had to be marketed. ERP will continue to provide a lot of fundamental data in the future, but the flexibility and clout you find in digital toolboxes such as Microsoft Azure are truly groundbreaking. The revolution in the planning profession has now really started. »
Bart : « One of the big differences compared to then, is that data management can be organized a lot more effectively today. Many initiatives from the past yielded only partial progress, because they often got stuck at the level of data management. Keeping master data synchronized across different systems proved to be very difficult in the past. Today's data lakes are smooth to use, making if really feasible to keep data consistent across systems, or even consider centralizing it. No small progress. »
Joël : « And since data integration is one of the major upsides of Cloud Computing platforms such as Azure… »
I can tell that there is no lack of passion. What are the next steps on your path?
Bart : « well, we really want to share that passion and so we have planned to organize a series of hands-on workshops. Specifically aimed at Supply Chain Professionals. »
Joël : « We need to tell people that this is the biggest revolution in our profession, since the arrival of ERP. We really want to explain Supply Chain professionals how this new technology stack can help them. »
Bart : « Everything starts with explaining a number of practical use cases in plain language. We really want to show people exactly what you can do in a Supply Chain context with AI and Machine Learning. Introducing them to Data Lakes or IoT. »
Joël : « And no: Python is not a beast to be afraid of. » (laughs)
Thank you for the interesting conversation. I am already looking forward to the series of workshops that you will be organizing. Good luck and thank you very much!
Are you interested in these webinars, don’t hesitate to send an e-mail to supplychain@moore.be and we will keep you informed.
