The car manufacturer announced a collaboration with Honeywell to find quantum answers to its real-world optimization problems.
This article was originally published on ZDNet.
BMW has become the latest company to take interest in the behavior of particles of matter taken at their smallest, quantum scale – or rather, in how those particles could generate the leaps in productivity that are expected to come with the advent of quantum-based technologies.
The German automotive giant is piloting the use of quantum computing tools to optimize the company’s supply chains for car manufacturing, and has now unveiled promising results from its early trials.
The quantum experiments were carried out as part of a new collaboration between BMW and US multinational Honeywell, which recently took its first steps on the quantum stage by making a trapped-ion quantum computer available to its customers over the cloud.
BMW is also working with a Singapore-based startup, Entropica Labs, which designs software that can be run on quantum computing platforms such as the one offered by Honeywell.
“The BMW Group is always exploring new technologies to further enhance our operations,” said Julius Marcea, head of IT at BMW Group. “We are excited to investigate the transformative potential of quantum computing on the automotive industry and are committed to extending the limits of engineering performance.”
Underpinning the trials is an effort to boost the efficiency of the car manufacturer’s supply chain, which is riddled with complex logistics to keep materials, good and services flowing smoothly between wholesalers, distributors, retailers, and ultimately customers.
To avoid as many glitches and disruptions as possible, it is necessary to make sure that the right products are always at the right place and at the right time; that’s a data-heavy task, stocked with ever-changing conditions and factors.
Reaching the most optimal configuration for supply chains is an equation that is often too difficult for classical computers to solve well, or in a reasonable amount of time. But quantum computers, which leverage the properties of quantum mechanics, are expected to be able to take on the most complex problems. They do so by running several calculations at once thanks to a special quantum state that is taken on by tiny particles inside the computer called qubits.
Building a quantum program requires a team that is qualified to develop the algorithms that can run on quantum computers, which is where Entropica Labs’ experts came in. The researchers analyzed the potential for quantum to play a role in BMW’s supply chain and assessed the performance of Honeywell’s H1 quantum system for the optimization problem at stake.
Entropica Labs’ team used a known quantum algorithm, called a recursive quantum approximate optimization algorithm (RQAOA), which is suited to the optimization problems that are central to logistics and supply chains. The RQAOA was then run on the H1 quantum system, a ten-qubit piece of hardware that was unveiled by Honeywell last year, and pitched as one of the most high-fidelity quantum technologies available today.
Given the small number of qubits currently available to run quantum programs, the technology was trialed on a small-scale problem that could also be solved by classical means.
“Since we only have small quantum computers at the moment, we are obviously not doing anything that couldn’t be achieved much more quickly and cheaply on a classical computer,” Ewan Munro, CTO of Entropica Labs, told ZDNet.
“The motivation is more to analyze the results from the quantum device and to use that information to help us build a strategy for moving forward,” he added.
The scientists were able to compare how the experiment performed compared to classical solutions, with seemingly promising results: Honeywell’s quantum hardware was shown to be competitive against a similar experiment run with a simulator, which uses classical devices to predict how qubits will react to different operations.
On the software side, the quantum algorithm also performed comparably to a classical algorithm called the Karmarkar-Karp heuristic. For the scientists, this suggests that once quantum systems gain enough qubits to support more complex problems, larger versions of the RQAOA could outperform leading classical algorithms.
BMW’s experiments with Entropica Labs on Honeywell’s hardware, therefore, rather constituted a proof-of-concept that was useful in validating the possibility of value being drawn from quantum technologies in the future. “The idea was to probe what quantum hardware can do today, and compare the results to what you would get using classical algorithms,” said Munro.
“The ultimate goal is to understand if and when we might reach a possible quantum advantage,” he continued. “We think we are still some years away from quantum advantage for real-world supply chain problems, primarily because it will take time for the hardware to mature to the required level.”
Honeywell’s ten-qubit H1 system only marks the start of the company’s pledge to launch additional generations of hardware with increased capability. It might be a few years, however, before quantum computing translates into observable business impact.
“Customers engage with us because they are doing proof-of-concept algorithms,” Tony Uttley, president of Honeywell quantum solutions, told ZDNet. “Nobody is talking about millions of qubits right now. We’re talking about tens of qubits. End-user organizations are getting the proof-of-concept in place right now to showcase what you can do with quantum computers today, and it sets them up to take advantage of them as they increase in capability.”
Another heavy investor in quantum technologies, IBM recently unveiled a roadmap towards a million-qubit quantum system. The company expects that it will be able to provide IBM customers with a thousand-qubit-strong computer by 2023, which could already see the start of value creation for some use cases.
Image via ZDNet
