The 32nd annual POMS Conference, held online by the Production and Operations Management Society, came to a close at the end of April. The event aims to highlight innovative and salient research that has been made in the field of logistics, distribution, sustainability, and supply chain management. This year, the conference hosted quite a few members of the WHU community, including Professor Arnd Huchzermeier, holder of the Chair of Production Management, research assistants Devadrita Nair and Kai Wendt, and doctoral student Daniel Hellwig.
The four of them presented their most recent research, some of which was first introduced in the previously published “Playing with DISASTER: A blockchain-enabled supply chain simulation platform for studying shortages and the competition for scarce resources.” Written by Hellwig, Wendt, Professor Volodymyr Babich of the McDonough School of Business of Georgetown University, and Professor Huchzermeier, this initial work analyzes the potential of distributed ledger technology (DLT) on supply chain performance. To facilitate this research, the team designed a simulation platform, cheekily named “DISASTER,” which can be used by instructors and students from all around the world.
The first presented paper, and spin-off from that original work, “A Behavioral Simulation of Blockchain-enabled Order History Sharing and the Bullwhip Effect,” focuses on how the disclosure of competitive information affects order inflation. Although managers may wish to obtain as much information as possible—especially concerning competitors—the team discovered that order inflation increases when more data (i.e., past order history) is shared. This, in turn, exacerbates the adverse consequences of the bullwhip effect, i.e., fluctuations in order amounts within a supply chain. This phenomenon has most recently been seen after the panicked purchasing sprints at the start of the COVID-19 pandemic.
In their second paper, “Simulation of Blockchain-enabled Market for Supplier Capacity Trading Among Retailers,” the quartet analyzes blockchain-enabled markets where retailers face random demands and varying valuations of goods. They discover and elaborate on novel trading strategies and determine that there are significant improvements in inventory efficiency when markets are enabled.
A final, separate work, co-authored by Devadrita Nair and Professor Huchzermeier, “Predictably Unpredictable: How Judgmental and Machine Learning Forecasts Complement Each Other,” takes a close look at the supply chain in action, i.e., as it pertains to seasonal products. Such products, especially highly innovative ones with short life cycles, leave behind little historical data, meaning that companies active in that retail area often have to make decisions based on purely judgmental forecasts. To combat the problem of bias in human judgment, the duo presented an integrated model that combines experts’ knowledge of the market with a machine learning algorithm’s ability to leverage historical information. Using real data from Canyon Bicycles GmbH, a manufacturer housed in Koblenz, Nair and Professor Huchzermeier achieved a significant reduction in forecast error over the traditional methods.
The research performed by WHU’s Chair of Production Management is making clear the fact that these novel approaches and technologies can help transform and revolutionize supply chain management as it is known today.