On June 4, 2020, we held our latest online Meetup with the theme “Artificial Intelligence for Supply Chain Resilience”, which raised a major issue:
How to deal with the unexpected?
The Covid-19 crisis has underlined the vital importance of Supply Chain resilience.
We continued to work after the hackathon on an easy to use dynamic simulation / optimization application interface. So we decided to continue our series of MeetUps, and this time, everything was in English!
We invited Matyas Kocsis, Head of Logistics for LafargeHolcim Eastern Europe, who presented how DCbrain helps us to manage the unexpected.
Finally, Professor Martin Christopher from Cranfield University, an expert in the field of Supply Chain Resilience for over 20 years, enlightened us on the future of the sector following the impact of Covid-19. This Meetup welcomed more than sixty people with many questions.
First of all, let’s start with a definition of the theme of our meetup:
A resilient supply chain can be defined as…
“a supply chain with the ability to quickly recover unexpected events impacting supply chain performance”.
The Hackathon project: Using AI to support the supply chain in the face of a global crisis
We created a dedicated CoVid-19 pandemic extension to our product in 48 hours.
The objective was twofold: to better manage the supply chain of essential products such as masks or medicines in order to reduce the number of deaths from the pandemic, and to anticipate the major impact that deconfinement will have on logistics at the European level in order to avoid a stalemate in the economic recovery.
DCbrain looked at the supply chain and warehouse occupancy rates, rail and road transport flows, production and delivery points. Thanks to our INeS technology and the expertise of our team, we proposed a solution to optimise them and prevent hazards such as border closures and increased transport costs.
Supply chains are the backbone of our economy. Yesterday, we all took it for granted that global supply chains functioned seamlessly.
Today we know that is not true. A global health crisis such as CoVid-19 creates uncertainty: borders may close or factories outside Europe may not deliver on time. When such a shock happens, chaos prevails. The supply of surgical masks and other medical supplies are perfect examples. This is why DCbrain participated in this hackathon, put its know-how at the disposal of the European Commission to hack this problem and propose an innovative solution, based on Artificial Intelligence, which will allow the Supply Chain sector to restart quickly, in the best possible conditions.
DCbrain, an efficient way to manage the unexpected in your supply chain?
LafargeHolcim needed a digital solution to optimize their network in Central Europe. The goal was to ensure fast, data-driven decision making in a complex environment with over 3,000 shipper customers and more than 70 different product types.
Several points had to be addressed and optimized such as :
- End-user experience (data visualization, rapid simulation of results…)
- A simplified process (coordination of the different countries and functions in parallel…)
- Profits (maximizing sales revenues)
- Serving a changing and growing market (factories sold, shortages due to raw materials and production, capacity constraints …)
For this reason, LafargeHolcim called upon DCbrain and its innovative INeS solution to meet these 4 different objectives. How to test the capabilities of DCBrain? – Execute a pilot project with clear objectives. We have therefore taken various steps to meet LafargeHolcim’s expectations:
- Creation of a digital duplicate of the physical network
- Creation of several real-life scenarios
- Comparing them with real management decisions
- Evaluate the results of the pilot
Thus, we were able to observe that Lafarge’s expectations were all addressed and that the implementation of our pilot on their network allowed us to respond successfully to the various issues.
Profits were increased (TCO objective achieved – reduction of 0.75%), the end-user experience improved (reduced calculation time), the process was simplified (easy modification of constraints) …
For LafargeHolcim, the implementation of AI on their network was a success at all levels. DCbrain allowed them to manage the unexpected in their supply chain while maximizing their profits.
Here are 2 questions that were asked by participants after Matyas’ intervention:
1) What kind of optimization algorithms are used in the demo?
We use a mix of classical algorithms such as Operations Research and more “modern approach” such as Reinforcement Learning. Our goal is to be able to quickly deliver a good quality solution.
2) What type of data do you use to be sure that the route we will choose is not as close (with weather risk, strikes, … for example) and will not ultimately increase the final cost? We consider that the operational team using the tool already knows which route is available today or not. Therefore, they can easily configure the tool to close warehouses or roads. Of course, unexpected events do occur. In this case, the operator marks the new road as closed (for example) and our tool is able to re-optimize very quickly to provide a new response. The objective is to provide a quick recovery solution.
What future for the supply chain after the COVID-19 crisis?
The supply chain, like many sectors, has been and will continue to be impacted by the current global crisis. However, the supply chain, as discussed in this meetup, will need to restart quickly as it is “the backbone of our economy”. For this, several long and short term plans will have to be put in place, in particular thanks to Artificial Intelligence.
In the short term, according to Professor Martin Christopher, several actions will have to be put in place such as :
- Establishing a multidisciplinary recovery team…
- Auditing the current supply chain to identify bottlenecks and timing issues
- Prioritize key customers/products
- Review the BOM for each product to identify substitution possibilities.
- We have to look at the long term with other action plans:
- Speeding up the digitisation programme
- Testing the supply chain – creating a “Digital Twin”.
- Look beyond Tier 1 suppliers to identify risk exposure
For these action plans to be implemented in order to “re-launch” the supply chain, the use of artificial intelligence will be necessary in order to improve network forecasting, to be able to cope with the various constraints, to optimise network flows, etc.
Thus, we can observe that real changes will be necessary to allow the network to restart properly, but this will of course be possible, especially thanks to AI.
“It is not the strongest of the species that survives, nor the most intelligent, but the most receptive to change.
Our thanks go to LafargeHolcim and Professor Martin Christopher.
This meeting will have allowed us to become aware of the importance of AI with different examples and use cases. Recovering from such a global crisis is not going to be easy. However, artificial intelligence is mature and ready to be used on all types of networks in order to optimize them and reach an optimality that makes us wonder about one point:
What are you waiting for to inject AI into your logistics network?
To go further, find our resources on the subject of Supply Chain and Artificial Intelligence:
Download our Logistics white paper: