Fujitsu Applies Quantum-Inspired Optimization Services in Supply Chain Logistics
Author’s Note: I do not have any relationship with any of the companies featured in this article that would create a conflict of interest.
On February 17, 2020, FreightWaves ran my column, Commentary: Toshiba’s simulated bifurcation machines may optimize your supply chain. Articles like that one often meet with skepticism from people with whom I often discuss developments in supply chain technology (#SupplyChainTech) — it just seems too “far out there” to be taken seriously by people who understand supply chain and know what’s happening.
So you can imagine my surprise to receive an email alerting me to a new development that builds on and confirms some of the assumptions I made while writing that article. On December 8, 2021, Fujitsu and the Hamburg Port Authority announced that they have successfully applied quantum-inspired algorithms to cut traffic jams and greenhouse gas emissions.
This is a big deal, because this problem is a real-world combinatorial optimization problem, and as I explained in the February 2020 article, “Combinatorial optimization problems are problems in which we seek optimal outcomes from a finite, but exponentially large set of possible solutions. Generally, this category of problems is solved using mathematical techniques, and the problem is governed by several constraints. Also, the set of possible solutions is too large for an exhaustive examination of each possible solution.” The types of combinatorial problems we encounter in the real world are extremely hard to solve.
If you have paid attention to the news recently, you know this is a meaningful development, given the inflationary pressures that the global economy is facing due to bottlenecks in global supply chains arising from the extreme mismatch in demand for consumer goods brought on by pandemic induced lockdowns, as well as the logjams that have occurred in ports in different parts of world. This development offers the potential that some ports, those that have invested sufficiently in information and communications technology infrastructure to be able to implement this new software, will be able to deploy one more tool in their efforts to ease the logjams that have plagued them for some months now.
According to the announcement, “The approach is a world-first for HPA, which is developing a new, holistic way to optimize all its available infrastructure of roads, crossings and bridges controlled by traffic lights. These are within an environment constrained by tight geographical limits, the cost of shipping schedule overruns, and specific maritime factors such as tides.”
Hermann D. Grünfeld, Head of Traffic Management at HPA, said: “With the global demand for transportation growing relentlessly year on year, logistics infrastructure owners and operators, like HPA, have to find solutions to the challenge of a finite physical footprint for most assets. It’s a common supply chain issue today and will get more important in the coming years. At the same time, we are seeking to reduce our carbon footprint as part of global efforts to address the climate emergency. Optimization is the logical way to address both these challenges. By working with Fujitsu’s quantum-inspired optimization services, we are delighted about the new capabilities to deliver greater capacity for our supply chain partners as well as lowering CO2 emissions.”
His counterpart, Dr. Joseph Reger, Chief Technology Officer (CTO) for Central & Eastern Europe at Fujitsu, commented: “Finding and delivering sustainable solutions for business and society is at the core of Fujitsu’s Purpose. Working with HPA, we have demonstrated the possibility and wider potential of applying quantum algorithms to the logistics industry. The business and environmental benefits we are developing with HPA show that Fujitsu’s quantum-inspired optimization services deliver results today — years, perhaps even a decade, ahead of the timetable for commercially usable quantum computers. These are improvements available today right across the Mobility sector — not just in maritime logistics.”
This milestone was reached by applying Fujitsu’s quantum-inspired Digital Annealer technology and services to real traffic data from HPA’s operations, enabling HPA to establish the possibility of optimizing traffic throughput across the harbor area, while still leveraging the existing traffic control infrastructure.
Fujitsu’s Digital Annealer is a computing architecture inspired by quantum phenomena. As I pointed out in my February 2020 article, such technology offers users the ability to rapidly solve complex combinatorial optimization problems at speeds significantly faster than general-purpose computers without the inherent complications typically associated with quantum computing.
The advantage of this approach is that it is possible to optimize the entire system at once, rather than the alternative of optimizing discrete subsections of the system. As Fujitsu and HPA say in their announcement, “Instead of local management of individual traffic-light managed crossings, the quantum-inspired approach optimizes the entire grid. This significantly cuts dwell times for ships, trucks and cars, resulting in faster supply chain interactions — and leads to lower levels of greenhouse gases.” Generally this approach is not possible on classical hardware so companies like Fujitsu and Toshiba that are working on real-world applications of quantum-inspired algorithms also have to develop proprietary hardware architecture to facilitate the implementation of these techniques.
In this case, the Graz University of Technology provided comprehensive support to Fujitsu and HPA throughout the process given its research focus on traffic planning and optimization.
The Fujitsu Digital Annealer performed real-time optimization calculations according to specified requirements and constraints, optimizing the network in less than 10 seconds. HPA believes this approach is scalable for its metropole network which has hundreds of light-controlled intersections.
HPA believes this approach will yield significant improvements in the following areas: Reduced travel time up to 15% for cars and trucks in the supply chain; Faster journeys to and from work for employees; Lower CO2 emissions from trucks and cars; Faster turnaround for container ships leading to increased flow of goods; Greater capacity to handle trucks in the confined space of the HPA, and; Less traffic congestion within the harbor area.
This development has happened much sooner than I expected. In the February 2020 column I observed that, “I am not suggesting that this particular technology will be available for use in the field tomorrow, or even this year. But I would be surprised if companies have not already started exploring how to use this new discovery in practical ways that go beyond a limited proof of concept.”
It will be interesting to see how long it takes before this technology is widely deployed in supply chain logistics networks all around us. For that, I think we’ll have to wait a while longer. But, I do not make predictions, so I am not going to hazard a guess when that will happen.
If you are a US-based startup working on making technology like this widely available to companies, and you are raising a pre-seed or seed round, come talk to us at REFASHIOND Ventures. We are finally in business, and #SupplyChainTech is the only thing we care about — basically, we’re the investor you’ve been waiting for all your life.
