Greater robotic precision and quantum-powered digital twins could generate better production performance, but there’s a catch…
Quantum computing technologies are advancing and will likely transform many industries over the next decade. There’s much focus on the emerging technology’s use cases in finance, cybersecurity, material sciences, pharmaceuticals, and high-energy physics. However, industrial automation is one of the many other areas where transformation is expected.
The greatest uncertainty about when the quantum revolution will occur lies in the challenge of increasing the number of qubits—also known as quantum bits— in gate-based quantum computers. The crux of problem is that gate-based quantum computers are currently very error prone, and although increasing the number of qubits they use increases their processing power, it also exacerbates the error issue.
While quantum physicists and engineers race to develop better error-correction techniques, enterprise early adopters are increasingly harnessing quantum computers made accessible via the public cloud to push performance gains in optimization problems, statistics and simulations while patiently working around the errors. Others embrace quantum annealers, which are devices that aren’t viewed in the same league as the holy grail of a fault-tolerant gate-based quantum computer, but are showing more immediate practicality as they aren’t as vulnerable to errors. There’s also hope that quantum analog simulators will show speedier progress compared to general-purpose digital quantum machines.
The growing subset of early adopters is expected to drive the global quantum computing market’s expansion by 497% between 2022 and 2028. Companies like BMW, Toyota, Hyundai, Airbus, Fujifilm and many others are already getting practical value out of quantum computers despite the sector not yet reaching the milestone of quantum advantage. Although there’s widespread uncertainty about when the threshold will be breached, IBM is one of several players that’s ramping up efforts, targeting quantum advantage by 2026 through quantum error correction (QEC) techniques.
A Hyperion Research study found that 80% of 300 organizations surveyed plan to increase their quantum computing investment over the next two to three years to enhance business efficiencies and increase revenue. The respondents included a wide range of use cases but the top ones were machine learning applications, finance optimization, logistics/supply chain management and modeling/simulation.
Part of what’s driving the appetite for quantum investment is the view among some technologists that classical machines are approaching the limits of Moore’s Law, which would slow future gains in processing power. For the area of industrial automation, quantum acceleration would mean that manufacturing operations could continue to grow in size and complexity well into the future. Because quantum computers make simultaneous calculations versus the sequential calculations of classical machines, they could enable factory robots to move with greater efficiency and precision, driving better throughput for more complicated tasks.
There are many ways in which quantum computing could impact manufacturing. Because the physical world is based on quantum mechanics, quantum computers are well-suited to simulating molecules and their interactions, which will advance the creation of new materials for use as semiconductors, industrial production catalysts, electronic components, sustainable fuels, pharmaceuticals and consumer products.
This could be especially helpful for automotive, aerospace, and electronics manufacturing by producing materials with better strength-to-weight ratios and batteries with higher energy densities, for example.
The automotive industry is already implementing quantum computing to solve manufacturing and design challenges with Volkswagen launching a quantum computing research team in 2016. In May, BMW and quantum startup Pasqal, headquartered near Paris, France, began collaborating on using quantum computational algorithms for metal forming applications modeling, expanding a previous partnership using quantum-inspired methods for computational chemistry to improve EV battery R&D. Quantum simulations also have applications for crash testing, parts design, and carbon emissions reduction.
The quantum approach using Pasqal’s neutral-atom quantum processors that use a digital-analog implementation of quantum algorithms, reportedly yields greater computational power to achieve digital twin simulations of more complex car parts with higher accuracy.
In April, Pasqal partnered with Siemens Digital Industries Software on a multi-year research collaboration to advance quantum simulation to aid in product design and testing for Siemens’ clients including Johnson & Johnson, LG, Airbus, BMW Group, EDF, Thales, MBDA, Credit Agricole CIB and others in the automotive, electronics, energy and aerospace sectors.
“Our collaboration with Siemens will explore how quantum computing can benefit Siemens’ customers who are looking for more accurate ‘digital-twin’ prototyping, which can reduce the need for costly and time-consuming physical prototyping in sectors like automotive, electronics, energy and aerospace,” said Georges-Olivier Reymond, CEO and founder of Pasqal.
Additionally, Multiverse Computing, a quantum software developer in San Sebastian, Spain, has partnered with the Bosch Automotive Electronics plant in Madrid to harness quantum computing to create a digital twin of a factory. Multiverse’s quantum-based optimization algorithms will assess the performance of equipment and production processes to enhance efficiency, energy and waste management.
Quantum computing is already enhancing supply chain management. The computational power of quantum annealers can provide optimal transport routes, enhance fleet efficiency and improve air traffic control. Quantum computing has also reportedly played a role in getting food to North American grocery stores more efficiently. According to a report from Deloitte, a Canadian grocery store chain used quantum computing to reduce optimization computing time from 25 hours to seconds. Additionally, the burgeoning field of quantum sensing is poised replace traditional field sensors to provide greater accuracy for industrial monitoring in factory settings or environmental monitoring in the field.
There’s still significant uncertainty about the breadth of quantum computing’s use cases. Their computational power has strong potential for some problems, but researchers anticipate that there will continue to be other problems that classical computers are better at even after quantum advantage is attained. Yet machine learning is integral to industrial automation by enabling computer vision, gaining data-driven insights about manufacturing techniques, and creating digital twin simulations, among other applications, and quantum computers hold promise for speeding up machine learning model training and running.
The convergence of advancement in the areas of AI, high-performance computing and quantum computing, could supercharge statistical computational ability for a broad range of applications, which bodes well for industrial automation and other fields. But for this to come to fruition, quantum and classical systems need to integrate, especially in environments like the factory floor filled with classical IoT devices. Quantum Machines, a company based in Tel Aviv, Israel, that’s developing a quantum computing platform, is focused on developing the quantum control layer, which includes all the hardware and software classical machines need to communicate with quantum hardware. According to Quantum Machines CTO Yonatan Cohen, the company has quantum control layer products that are compatible with almost any type of quantum hardware currently deployed to over 200 customers globally.
“Quantum hardware is not where we want it to be—it’s still at the research and development stage,” said Cohen. “In order to achieve practical applications you have to really squeeze the quantum hardware for every bit of performance. One of the best ways to do it is by offloading as much as you can from the quantum hardware to the classical hardware and only use the quantum processor for very specific things that it can do well. People are coming up with more creative algorithms and protocols to do so, but we need the right integration in the hardware and underlying software.”
One area where quantum computers are showing the most practical advantage is in cybersecurity, which also has implications for industrial automation. Quantum computing companies are offering some of the strongest encryption methods available such as Quantinuum’s provable, patternless or non-deterministic entropy that it’s inserting into encryption keys for clients. With cyber attacks increasing globally by 125% in 2021, it’s essential that industrial operations such as manufacturing and energy invest in the leading cybersecurity frameworks to stay ahead of the threat.
Yet quantum’s safety provisions could quickly morph into a threat source. That’s because as the technology advances, it’s expected that a quantum computer will eventually emerge that’s powerful enough to run Shor’s algorithm and break classical encryption in an event called Y2Q —a global catastrophe. But according to Zapata Computing CEO Christopher Savoie, quantum’s threat to classical encryption could arrive long before a computer capable of running Shor’s algorithm emerges, explaining his team was able develop a heuristic algorithm, called variational quantum factoring (VQF), in the course of a single weekend that could compromise some instances of RSA encryption using noisy intermediate-scale quantum (NISQ) devices in the near-term. By extrapolating their results, the team estimated that with just a little over 6,000 NISQ qubits, VQF could factor a 2048-bit RSA number in a few hours.
“If you’re a bank, energy company or government, that’s not an acceptable risk,” said Savoie who recommended that the US National Institute of Standards and Technology study the issue as part of its post-quantum cryptography strategy and that enterprises adopt a more crypto-agile approach.
The vulnerability could be significant in the industrial automation arena with IoT devices connecting networks of factories that could be vulnerable to quantum-driven DDoS attacks, taking operations offline. The attacks could also expose sensitive enterprise Intellectual Property data.
According to Sandbox AQ VP of Product Nadia Carlsten, cyber-physical networks such as factory robots and other IoT devices can be more challenging to secure as algorithms or keys that are long or require a higher level of processing can create detrimental performance issues. Although a crypto-agile approach could transcend these issues in the short-term, in the long-run quantum computers might be both a boon and a pitfall for industrial automation as the sector experiences quantum productivity gains, meanwhile the Y2Q threat approaches.