Launched in fall 2020, the Extensiv Supply Chain Scholarship aims to foster the talent that will drive the future of our industry, inspiring fresh ideas and voices in logistics and the supply chain.
Below, we present Rena Ben-Eliezer, a runner-up in the Fall 2024 Extensiv Supply Chain Scholarship, and their essay on how quantum computing—based on the principles of quantum physics—will solve the most troublesome and complex problems that have perplexed the logistics and supply chain industry for years, especially route optimization.
Seconds, Cents, and Organ Transport: Imagining a World with Quantum Route Optimization
This is not Sasha’s first time in surgery. Since age three and her diagnosis with a rare liver cancer, her world has been an Escher staircase of therapies, drugs, and experimental interventions. Favorite doctors replace favorite teachers as her childhood heroes. Her earliest memories are riding the elevator in glass-walled medical centers.
In the lobby of the Cleveland Clinic, where Sasha awaits her liver transplant surgery, sits one of the newest quantum computers—a small Tardis-looking machine. It was installed in 2023 in a flagship joint-venture between IBM and the Cleaveland Clinic to accelerate biomedical research. Although the clinic’s director, Dr. Morton L. Mandel, is candid about the current barriers to curing diseases like Sasha’s, he says the team anticipates “potentially find new treatments for patients with diseases like cancer, Alzheimer’s, and diabetes.”1
How does it work?
Quantum physics, suggested roughly one hundred years ago in the seminal papers of Niels Bohr and Max Planck, is only just beginning to be harnessed by industry. Pilot projects are underway in fields as diverse as chip fabrication and aerospace. While classical computing relies on the binary values of matrices of transistors, a quantum computer’s “qubits” can hold multiple values simultaneously—called superposition—and become entangled in near infinite permutations. This enables the rapid solving of problems so complex that they overwhelm even the most advanced deep learning algorithms.
The traveling salesperson is not “a problem, it’s an addiction,” according to computational complexity researcher Christos Papadimitriou.2 It is the earworm of computer science graduate students. Deceptively simple, the problem seeks the shortest or most cost-effective route through a dispersion of nodes, which can represent houses waiting for packages, oil wells that require servicing, or circuit board transistors that need soldering. A problem containing just 15 nodes has 87 billion independent route choices. Yet, even rudimentary quantum systems find answers in seconds that took traditional processing units days to compute, a process dubbed quantum annealing.
Consultants have projected the firm-level impact of this breakthrough. Academic publications speculate on emissions and fuel savings. However, few address the looming macroeconomic implications. Each year, economists earmark roughly 10% of GDP for logistics activities—a $10 trillion dollar industry set to double by 2030.3 In economic terms, this spending mostly falls under transactional costs, the unavoidable inefficiencies of goods and dollars changing hands within a market. Heuristic algorithms used for route optimization today can consistently generate routes 150% costlier than optimal. With this in mind, the implications of a 30% more efficient logistics system grounded in quantum annealing are hard to overstate.4
First, businesses can take advantage of previously unprofitable exchanges. According to most trade models, if transaction costs dip below the marginal gain of trade, that trade now becomes viable even if only a few cents are made per transaction. In a post-pandemic world where shoppers are particularly price sensitive, goods will travel farther to retail shelves where they can be a few cents cheaper—for example, the grocery chain Carrefour’s refusal to carry Pepsi products earlier this year. Importation of fresh vegetables has become increasingly common.5 In fact, Aldi and Walmart are already adopting the strategy as a hedge against inflation.
Second, lower transaction costs inevitably lead to increased competition. What the invention of the container did for Global shipping in the 1990’s, quantum route optimization will do for last-mile logistics today. A customer placing an order on Amazon is typically presented with between one and five different shipping options and prices. The Amazon checkout screen of the future could include 50 options, with multiple LTL and parcel providers running quantum annealing algorithms in real time to find the marginal cost of including that consumer’s front porch in their routes. The result is likely to be less costly and more efficient than ever before for consumers.
On the firm side, of course price is only one consideration. Companies reward responsiveness, fill rate, lead time, and lead time variance—and quantum optimization can address these variables as well. It’s easy to imagine logistics providers including their client’s inventory demands and economic order quantities in the route costing equations. This will reduce lead time variance, by extension lowering reorder points. Hot shot and milk runs will become inexpensive and unnecessary, dropping buffer stock levels. In short, products will spend less time on the shelf and more time in transit, lowering holding costs. The supply chain will become a supply pipeline.
So, better than a cure for cancer?
From a perspective of net well-being of most populations, healthcare is still largely an issue of access rather than capability. In the global south and in remote areas, the problems become particularly dire. For instance, during the COVID pandemic, rural states like West Virginia struggled to access PPE (personal protective equipment) and vaccines. In extremely volatile situations, optimal routing becomes more important than ever.
However, optimized routing can only optimize so far. The complexity stems from assigning costs to different outcomes. In its base form, the traveling salesperson problem looks only to minimize costs between nodes. We have discussed models where different nodes are weighted according to client lead times, but how do we weight the less tangible cost of delivering a lifesaving vaccine or a hospital experiencing stockouts of PPE? With a plurality of perspectives and incomplete information, these are difficult decisions to build consensus on, let alone communicate to an annealing algorithm.
Sasha’s liver donor lives over 2,000 miles away in rural Baja, Mexico. Cell death occurs within 24 hours. What is it worth to her, to her care providers, to the American economy, to get it to her on time?
Sources
1. Cleveland Clinic. (2023, March 20). Cleveland Clinic and IBM Unveil First Quantum Computer Dedicated to Healthcare Research. Cleveland Clinic. LINK
2. Klarreich, E. (2013, January 30). Computer Scientists Find New Shortcuts for Infamous Traveling Salesman Problem. Wired. LINK
3. Maiden, T. (2020, January 11). How big is the logistics industry?. FreightWaves. LINK
4. Klarreich, E. (2020, October 8). Computer Scientists Break Traveling Salesperson Record. Quanta Magazine. LINK
5. Kriege, M. (2023, October 5). Declining demand for perishable goods: how must supply chains adapt?. Maersk. LINK
