Step-by-step plan optimises forecasting in logistics chains with AI
Can artificial intelligence (AI) be used to better predict and plan supply chains? And can it help logistics chains maintain efficiency in the face of unexpected events? The answer is yes, according to the VIL project ‘AI4Cast’. Using AI for optimal inventory planning can even deliver cost savings of 25% or more. VIL provides guidance to companies that want to make the leap.
Supply chains are becoming increasingly complex, making it difficult to plan and predict demand and capacity. Unexpected events also complicate matters, such as unexpected spikes in purchasing behaviour due to hype, delays due to unforeseen weather conditions or material shortages due to the pandemic. Such events often trigger a chain reaction, causing issues for multiple companies. This leads to costs such as additional transport, overtime, or space or equipment hire.
Intuition has long since become outdated. For companies to be able to work at optimum efficiency, supply chain forecasts and planning must be as accurate as possible, whatever the circumstances. All companies in the chain – as well as the end client – benefit from real-time, proactive forecasting across the entire supply chain.
VIL’s AI4Cast project shows that AI provides an opportunity to handle supply chain planning and forecasting more efficiently. Other conclusions include:
- High-quality data and a willingness to work together are crucial for the successful use of AI applications
- In an ideal world, forecasting would not be used unilaterally – there should be collaboration between every company that forms part of the same supply chain
- Data standardisation is extremely important
- Algorithms are only as good as the data fed into them. The right AI technologies applied to complete and valuable data will minimise the margin of error in planning and forecasting
- As well as greater efficiency, algorithms also deliver cost savings: depending on the quality of the data, the business processes and the degree of collaboration, they can result in cost savings of 25%, or even more.
These conclusions were able to be made because of pilots at Colruyt and Procter & Gamble. This case study focused on demand forecasting of a number of categories of goods purchased by Colruyt from Procter & Gamble.
Based on the pilots and the conclusions, VIL has developed a step-by-step plan to optimise companies’ forecasting with AI. Companies can use it to determine what data is available; how they can use intelligent algorithms to generate valuable information from this data; and what collaboration with the logistics partners concerned – with the help of a reliable data platform – is appropriate.
Guidance is important
“Not all companies are sufficiently mature when it comes to data,” says project leader Dirk Jocquet. “This means application of the step-by-step plan might not come naturally. What progress have they made in collecting data? Is it the right data? Do they have suitable systems to process the data? VIL can guide them through this analysis. Companies can only move on to the next step once they have their data in order.”
If you’re interested in having VIL guide you through analysing the data and implementing the step-by-step plan, contact Dirk Jocquet: email@example.com