Algorithms as a Tool for Integrating Supply Chains
Posted by Louis D'hondt on 06/12/2017
A close collaboration, a merger, an acquisition or even the start-up of a new business unit. Such interventions pose major challenges to the supply chain. Paying too little attention to the planning process is a classic trap. We examine how algorithms can help.
“From the data, you have to move to pragmatism and reality…This pragmatism can be very specific and can be incorporated into your algorithm, which humanises data-driven work.”
Business Unit Manger & Strategic Information Manager,
Comparing and integrating the methods and the data of two organisations. In theory, merging supply chains seems clear enough, but in practice a supply chain assessment after a merger almost always surfaces unexpected challenges. Addressing these is one thing; subsequently also taking advantage of the opportunities of the merger is another. Here, a strong algorithm can provide many useful simulations. But how to translate this into a focused plan?
Integrate without sacrificing internal dynamics
Integration is obviously the core task of a supply chain merger, but many companies attach importance to healthy competition between different business units or hubs. Johan Ghijselinck is manager of the General Cargo business unit at H.Essers. He is responsible for a revenue of 115 million euro, he operates a fleet of 1 000 trailers, 400 trucks and 500 FTE's.
The company in total generates a revenue of 571 million euro, operates a fleet of 3 050 trailers, 1 360 trucks, and manages 5 400 FTEs. At the same time, as Strategic Information Manager, he is responsible for merger and acquisition projects. “Integration is the message, but if you make an acquisition or start a new branch, such a business unit must retain some self-sufficiency within the organisation”, he confirms. “Then you are indeed organising a kind of internal competition. In line with this, the commercial organisation is local and our simulation tools provide analysis. Suppose that a German branch picks up a new customer from the pharmaceutical industry. It’s possible that the tool shows that the customer would best be handled in Genk. But that doesn’t detract from the fact that the intention is for the German branch to also generate business.”
Think through change management
Yet Johan Ghijselinck wishes to immediately state that bringing together supply chains is more than blending processes. It’s about people. Change management must focus on this to a significant extent and not just on the classics, such as involvement in the change. “First, in pursuing such an objective, you need the support of the operational manager. Such hierarchical support is extremely important.” But even more crucial is coordination with the planners. “They must have good insight into how much knowledge they possess and how crucial they are to making the data ‘planable’. It might be advisable in this context to invest in information systems. They make it possible to better bring this internal knowledge to the surface.”
Create clarity early on
Such change management inevitably leads the organisation to the point where people wonder what their future role looks like. These are questions that need to be answered at the earliest possible stage. At H.Essers, the goal is to grow with the same number of people. “It’s imperative that you clarify what the consequences are of the roll-out of systems that partly take over the work of planners”, says Johan Ghijselinck. “Think in advance about possible new tasks or about who you will have to dismiss. If you don’t include this in the exercise from the beginning, people will form their own opinion, which is less efficient in terms of strategy.”
From algorithm to pragmatics
However, merging supply chains does not only mean re-examining the task of people. It is also very practical to clearly define how you want to deal with buildings and vehicles for example. “We do this in very data-driven way”, says Johan Ghijselinck. “We apply the algorithm to this aspect with a focus on the number of kilometres and the working hours of the drivers. Based on the reductions here, it becomes possible to estimate why, for example, you will no longer need to use a hub.” And that doesn’t have to be based on complicated calculations. Johan Ghijselinck argues for pragmatism at several levels. “From the data, you have to move to pragmatism and reality. Some partners, just to give an example, might no longer be required for export, but are useful for import or for another hub. This pragmatism can be very specific and can also be incorporated into your algorithm. Take efficiency levels for example. A driver who drives to Tienen every day and suddenly has to go to Vilvoorde is working less efficiently than when working in Tienen. This can be built into the algorithm, which humanises data-driven work.”
Looking for the opportunities
There is also the question of how in this process to also keep the focus on the opportunities offered by the merging of supply chains. Johan Ghijselinck gives the example of acquiring cross-dock warehouses. “It’s important to find out exactly where your customers and loading points will be consolidated”, he says. “For some industries, such as pharmaceuticals, that’s an important question because they require that a shipment only hit one cross-dock. Conversely, if a shipment may hit multiple cross-docks, you can use algorithms to find out where the potential for savings lies.” This is the type of insight that is making logistics players adopt a new position today. Simulation tools show the savings potential for the logistics partner and its customers. “The data from the algorithms encourage customers to see for themselves how they can optimise. It’s a bit of logistics consultancy: addressing the ‘what if’ questions, which makes you more effective”, concludes Johan Ghijselinck.