Building the self-driving supply chain

A long-established pharmaceutical player aims to become a leader in supply chain technology.

ALESSANDRO DE LUCA, Chief Information Officer, Merck Healthcare at Merck Group

The year 1668 was an important one for far-sighted individuals. In London, architect Christopher Wren was trying to persuade King Charles II to adopt his plans for the reconstruction of the city, much of which had been destroyed by fire two years earlier. Isaac Newton built his first reflecting telescope. And in Germany, when Friedrich Jacob Merck received a license to open a pharmacy in the city of Darmstadt, he founded an organization that is now the oldest chemical and pharmaceutical company in the world.

Today, 350 years and 13 generations later, the Merck family is still the majority owner of the company that bears the founder's name. Over that time, Merck has been through many transformations. It has grown from a local pharmacy into a global manufacturer of chemicals and pharmaceutical products. It has acquired other major businesses and divested parts of itself.

Merck Group now employs 52,000 people in 66 countries. Its operations are divided into three divisions, each addressing different markets. Performance Materials is a leading producer of liquid crystals and organic light emitting diode (OLED) materials for the electronics industry. Merck's Life Science division, meanwhile, creates tools and materials for the scientific research community.

Alessandro De Luca is Chief Information Officer (CIO) of the company's Healthcare division, which produces a broad range of prescription and over-the-counter medicines and supplements, including cancer drugs, fertility treatments and products for the treatment of multiple sclerosis. He has been in his current role for 18 months, and ran the division's supply chain function for six years before that.

A digital transformation

Today, De Luca is spearheading a new transformation in Merck Healthcare. Its aim, he says, is to "leverage digital solutions to improve the life of the patient." Given his background, it is no surprise that De Luca sees digitalization of the supply chain as a central part of that effort. "For us, supply chain digitalization has two elements," he says. "The primary objective is to gain real-time information on customer demand, and to be able to predict future demand so we can produce what patients need in a proactive manner. Our goal is 100 percent on-time delivery of all our products. And secondly, we want to maximize the overall productivity of our supply chain."

The vision, he says, is the creation of a "self-driving supply chain," in which production and distribution adapt automatically to meet changing demand. That's an ambitious plan, however, and De Luca emphasizes that Merck is taking a phased approach to its digitalization plan.

"Our first step, starting about two years ago, was the creation of a supply chain control tower," he says. The control tower brought all the division's supply chain data together in a single digital dashboard, providing a centralized, real-time view of its end-to-end supply chain operations. The project was "very successful," De Luca says, helping the company to boost it service levels "from very good to almost perfect."

A year ago, Merck embarked on the second phase of its digitalization effort. This focused on improving the accuracy of its forecasts through the introduction of machine learning algorithms to automate the demand forecasting process. The impact of that effort has also been significant. De Luca notes that, while the overall accuracy of the automated forecasting system is similar to its previous processes, forecast variability has been cut by more than a third.

That matters, because in the healthcare business spiky demand is expensive. High service level requirements mean companies like Merck must manufacture and store costly additional inventory to cover likely peaks in demand. The smoother forecasts generated by its new system have allowed it to cut overall inventory levels by between 5 and 10 percent, with no negative impact on service.

Testing the autopilot

Work is now underway to make the vision of a truly self-driving supply chain a reality. "Forecasting is nearly all automated right now, and we are making good progress in the introduction of digital tools to automate our supply planning," says De Luca. "Now we are piloting the use of artificial intelligence to balance supply and demand." Eventually, he says, the AI system will be smart enough to recommend adjustments to purchasing, production and distribution plans, or even to execute those changes automatically without human intervention.

For now, however, that level of autonomy remains an ambition. Merck plans to pilot a fully integrated forecasting and supply planning process later this year, something that De Luca believes will be a first for the life sciences sector.

Being at the cutting edge of digitalization brings its own challenges, he adds. And the challenges are changing with each step the company takes. "When we built the supply chain tower, the technology was the easy part. There were several good options already on the market," he says. "The main difficulties came from ensuring that our master data was all correct and consistent."

As Merck has brought greater automation into its forecasting and planning processes, the human element has become more significant, says De Luca. "There is a significant change management challenge. It's about how people embrace a supply chain that is driven by AI. Are they willing to rely on a machine for business-critical decisions?"

Merck has addressed change management systematically. "We made it our policy that all digitalization projects are driven by business requirements, not just because the technology is there," he says. "And we've tried to learn fast, with lots of pilot projects, early failures and quick restarts." There has also been a significant communications effort to ensure people across the business understand the goals of its digitalization efforts, and its progress toward those goals.

Getting the technology right

BRIGHT IDEA: Merck is a leading producer of organic light-emitting diodes.

In the current phase of its program, De Luca notes that the technology itself has become a more significant challenge. Merck is partnering with specialist AI companies, including Palantir and Aera Technology, to develop its integrated supply chain management systems, but the technology to do everything it wants to do is "not yet there," he notes. "We are not far off, however, and I believe there will be a real first-mover advantage for the companies that do manage to integrate artificial intelligence into their supply chains successfully.

Ultimately, De Luca thinks artificial intelligence in the healthcare supply chain won't just be a benefit, it will become a necessity. "The whole world is going to be changed by digitalization. We will see more people using digital tools and deeper integration of sensors everywhere. In healthcare, that's going trigger a shift to patient-driven demand. Rather than the hospital or the pharmacy, you will have the patient at the center of the supply chain."

Creating systems that can predict and respond to demand from millions of individual patients will mean a step change in complexity, says De Luca, and it will bring new challenges in a host of areas, from technology to regulations and the control of data privacy. It will also require healthcare companies like Merck to bring unprecedented levels of agility and flexibility to their supply chains. "Patient expectations are going to be very high, and it will be up to us to create the supply chains that can meet those expectations." — Jonathan Ward



Published: June 2018



Images: Merck; Krisztian Bocsi/Bloomberg/Getty Images