HOPE (Human-Centered Optimization for Supply Chain Excellence)

A key goal of the topsector logistics is to “grow the added value of supply chain management (SCM) to 29 billion euros” while “reducing greenhouse gas (GHG) emissions”. However, tight labor markets have resulted in a “Big Supply Chain Talent Shortage” (Lebovitz, 2021) at a time when the need for SCM professionals has been exacerbated by increasing complexity in SCM because of increasing ambitions and regulation surrounding GHG, circularity, or trade restrictions: the Dutch logistics sector may lose its competitive edge unless this issue is resolved.

We present HOPE (Human-centered OPtimization for supply chain Excellence), an applied research project that studies and improves human-algorithm collaboration to unlock the potential of automated supply chain decision making based on optimization and artificial intelligence (AI). This can substantially reduce the demand placed on human workers: research has demonstrated that algorithms are capable of handling complex decision making contexts while increasing resilience, sustainability and performance in transport, warehousing, and production supply chains.

However, most algorithms have been designed with the intention to fully replace human decision-makers: real-world SCM applications almost always incorporate human oversight to integrate tacit or “soft” knowledge. AI systems designed with the perspective of full automation lack the design for intuitive human interaction, and it is increasingly recognized that the promise of supply chain automation can only bear fruit if algorithms are integrated based on a deep understanding of the human planners that will work with them; we need “a well-functioning socio-technical system where human-centered technology is implemented” (Nederlandse Kennisagenda Logistiek). However, the transition to such human-centric systems is complex, and hampered by a poor understanding of the conditions for effective collaboration between human and algorithms in realistic supply chain contexts, and a lack of techniques that can take the perspective of humans into account and that can help them in interacting with the algorithm.

The HOPE project works towards systems that seamlessly support human decision makers based on AI support, reducing the burden placed on supply chain professionals while efficiently dealing with the objectives arising from today’s supply chain complexities. To this end, we

1. develop novel algorithms and methods that underlie collaboration mechanisms to enable us to streamline human-algorithm interaction, building onto expertise in smart automation and drawing from experience of consortium companies and insights gained from experiments with human subjects.
2. perform first-of-a-kind experiments involving human subjects interacting with a prototype “supply chain copilot”: A task environment that enables humans to oversee an algorithm-supported decision making process for a complex, realistic supply chain, while being supported by various collaboration mechanisms. This yields key insights into how collaboration mechanisms, algorithms, training of the employee, and decision making environment promote or inhibit effective human-algorithm collaboration in a realistic supply chain setting.
3. integrate our supply chain copilot into workshops to create a full immersion experience for participating supply chain leaders, showcasing prototype human-centric collaboration in a relevant SCM environment, to foster broad awareness of the importance of considering the human while developing systems and tangible methods for achieving that.
   Our diverse team of researchers boasts technical expertise that spans several disciplines: behavioral experimentation with human subjects, smart automation, and stochastic optimization and AI for planning. Our consortium further comprises a committed group of seven leading companies whose willingness to contribute substantially to this project demonstrates their eagerness to take up and further develop the human-centered technology prototyped and tested in this project.

HOPE paves the path towards human-centric implementation of AI and automated SC processes in supply chain copilots, ensuring meaningful and efficient worker involvement, thereby allowing for an efficient use of the labor force and catalyzing economic growth in an era where labor shortages increasingly threaten to strangle economic growth and broad wealth.