Quantitative supply chain redefines how supply chains can be optimized with more capable software, typically driven by machine learning, and with more scalable software, powered by a Big Data infrastructure. However, at the core of any quantitative supply chain project, there is the supply chain scientist who executes the data preparation, the economic modeling and the KPI reporting. The supply chain scientist delivers human intelligence magnified through machine intelligence. The smart automation of the supply chain decisions is the end product of the work done by the supply chain scientist.
Human plus Machine
Improving the supply chain performance requires a deep understanding of the business strategy. Stock-outs can be dramatically expensive, as is the case in aerospace for example, or just business as usual, as it happens with fresh food. Realistically, today, while smart algorithms can win against chess or Go champions, even the smartest machines are decades away from being able to establish a strategic roadmap for your company’s supply chain. Thus, establishing a purely machine-driven set-up to drive your supply chain remains science-fiction.
However, smart algorithms and machine learning algorithms have become incredibly good at solving problems that are well-defined, narrow and repetitive. Quantitative supply chain embraces these modern software capabilities: it uses human intelligence to frame the problem, eliminate the ambiguities, and set up the repeatable workflow; and then, it lets the machine take over when it comes to generating the extensive, but mundane, supply chains decisions that your company requires every day to keep operating.
Quantitative supply chain is not about removing human insights from the picture. Quite the opposite actually. Quantitative supply chain is about bringing human insights back to where those insights have the most significant impact: strategic matters. It is precisely by liberating the supply chain staff from the mundane and repetitive tasks that quantitative supply chain gives back the freedom required by teams to concentrate their efforts on the strategic issues, instead of being stuck with operational details.
The role of the supply chain scientist
The role of the supply chain scientist is to “crunch” the data, to factor all the economic variables into the logic and to automate the generation of supply chain decisions. The supply chain scientist is also responsible for implementing and monitoring the KPIs, devised together with the supply chain management, that are used to assess the performance of the quantitative supply chain initiative itself.
At the beginning of the initiative, during the scoping phase, the supply chain scientist is responsible for making sure that the problem to be solved is well-defined, that ambiguities are, if not yet resolved, at least clearly identified as such. In particular, the supply chain scientist is responsible for establishing a clear picture of the intended automation. Depending on the context, the automation might be intended to generate purchase orders, inventory movements, inventory scraps, etc.
During the data preparation phase, the supply chain scientist must make sure that all the relevant data is properly extracted from the IT systems of the company. While the supply chain scientist typically gets some help from the IT staff to execute the data extraction itself, the supply chain scientist is the one responsible for making sense of the data. Establishing the precise semantic of the data, from a supply chain perspective, is of critical importance. Turning raw system data into prepared data that is ready to be processed by a machine learning algorithm requires a significant amount of effort. This responsibility falls on the supply chain scientist again.
During the onboarding phase, the figures produced by the automation are challenged by the supply chain practitioners. In this phase, practitioners often uncover edge cases when the automation misbehaves. And it is the supply chain scientist’s responsibility to fix those edge cases. However, it also happens that the “odd” numbers are actually the correct numbers, but diverge from past non-optimal habits of the supply chain practitioners. The supply chain scientist has the responsibility of shedding light on such situations and convincing the supply chain practitioners that those numbers do constitute a problem, but rather happen to be a key ingredient of the solution.
Finally, once the solution is in production, the supply chain scientist monitors the performance of the automation and identifies its weaknesses. The supply chain scientist is responsible for the continuous improvement of the solution. Frequently, an improved logic requires better or more data, which in turn requires changes in the operational supply chain processes. The supply chain scientist quantifies the expected gains associated with data improvement and builds specific business cases to propose the change to supply chain management.
The skills of the supply chain scientist
The supply chain scientist is both a data scientist and a supply chain expert. This dual competency is essential in order to succeed in delivering a solution that lives up to the initial expectations. Supply chain expertise is essential in order to make sure that the supply chain scientist has a deep understanding of the challenges that need to be addressed. A lack of understanding of the supply chain challenges puts the project at the risk of getting a “solution” that isn’t aligned with the supply chain needs. Varying lead times, MOQs (minimum order quantities), costs of air transport vs sea transport, multi-echelon analysis … are just some of the many angles that need to be mastered by the supply chain scientist. More specifically, fulfilling the role of the supply chain scientist requires a deep understanding not only of the elements themselves, but also of the relationships between them. For example: how the MOQs are influencing the lead times.
The data science expertise is essential in order to, first, perform quantitative assessments that leverage the historical data, and second, implement the logic that entirely automates the mundane decision-making process. A lack of programming fluency puts any initiative at risk of excessive delays and hazardous numerical results. Programming is a skill as well as an art. Supply chain challenges are incredibly complicated. The supply chain scientist is capable of implementing a solution that is simple enough to be sustainable, yet accurate enough for delivering the desired supply chain performance.
Finally, the supply chain scientist role also requires above-average communication skills. Good writing skills are important in order to produce high-quality documentation describing the quantitative supply chain initiative itself. In fact, supply chains are all about trade-offs - for example, smaller MOQs vs lower purchase prices - and too frequently those trade-offs tend to remain undocumented for the most part. Quantitative supply chain requires those trade-offs to be documented and quantified. The responsibility for this task falls on the supply chain scientist. Good verbal skills are required to order to engage in constructive dialogue with supply chain practitioners during the onboarding phase, especially since the supply chain teams need to be convinced on the validity of the new approach.
Supply chain scientists at Lokad
At Lokad, the supply chain science competency has gradually emerged over the last decade (Lokad was founded in 2008). While Lokad did start off as a pure software company, we came to realize that supply chain excellency required having a dedicated Lokad team to take action on the front lines when dealing with actual supply chain challenges. The traditional “software support” staff wasn’t anywhere near enough to bring satisfactory solutions to companies as this requires a deep understanding of many different supply chain challenges, not merely a deep understanding of Lokad’s technology.
Establishing and growing a supply chain science competency is difficult. As a result, many companies rely on Lokad to fulfill the supply chain scientist’s role for their own quantitative supply chain initiative. In this instance, Lokad provides a software+expert solution, where a supply chain scientist gets assigned on the case and start orchestrating the whole initiative. This approach relieves companies from the need to immediately establishing their own supply chain science competency. Outsourcing this competency makes sense for both small and large companies. For small companies, the costs of doing this in-house are just too great. For large companies, it is primarily a matter of accelerating the pace of change within their supply chain.
The types of candidates selected by Lokad for its supply chain science teams are usually engineering profiles with Master degrees. Although Lokad’s supply chain scientists are familiar with programming, they generally aren’t software developers. Instead, their skill mix tends to be more varied and includes most engineering fundamentals: the capacity to model industrial problems, to establish a process, to make this process both performant and reliable, to communicate with management, etc. Due to the very nature of Lokad’s supply chain challenges, we are inclined to select profiles that are fluent in mathematics and statistics given that these two fields are essential for the quantitative resolutions of most supply chain challenges.
The development of this supply chain science competency is an ever-improving process at Lokad. And since Lokad is fulfilling the supply chain scientist role for many companies across different verticals, we have been built significant institutional knowledge in this area. What’s more, when new recruits join Lokad, their training involves being exposed to diverse supply chain situations, in multiple verticals, in order to accelerate the learning process and achieve deeper levels of understanding.