Innovation is essential for sustained success in manufacturing. This means looking at new and emerging technologies to determine how, if at all, they can help boost efficiency and profitability and enable production of new products.
One of these emerging technologies is quantum computing. A big advance over classical, binary computing, this offers the prospect of solving hitherto insoluble problems. This blog looks at the potential impact of quantum computing in manufacturing, addressing applications, challenges and its relationship to artificial intelligence (AI). First though, we begin by explaining what’s different about quantum computing.
What is quantum computing?
At their core, the computers we use every day are still simple devices. Each transistor on a chip works like a switch and is either on or off. Those states are converted to 0 and 1, which form the basis of binary code that underpins almost all computer hardware and software.
A quantum computer applies the principles of quantum mechanics, which lets each computing “bit” have states between 0 and 1. This leads to a more probabilistic form of computing, capable of performing calculations many times faster than even the fastest of today’s supercomputers. These quantum computers can solve problems too complex for so-called classic computing and greatly compress the time needed to solve very large problems that would otherwise need months or years of processing time.
For many years, quantum computing was little more than a theoretical idea. Today, quantum computers and computing services are available from companies such as IBM, Google and D-Wave. The computers themselves are extremely complex and expensive to buy and run. Furthermore, quantum computing requires highly specialized algorithms and are known to be error prone.
So far, two major classes of application have emerged: materials science and chemistry, and cryptography. In the former, quantum computing is helping researchers find and evaluate new compounds, while in the latter, it threatens to render existing cybersecurity defenses obsolete.
Where it isn’t yet being used outside of research centers is manufacturing, but that may be about to change.
Applications of quantum computing in manufacturing
Maximizing efficiency in manufacturing requires optimizing complex, multivariable problems. Until recently, the problem was lack of data, but going forward, the Industry 4.0 revolution means the challenge is becoming how to use the data. Tools such as simulation and digital twins help, but sometimes the problems are just too large and too complicated for classical computers to solve in a reasonable timescale.
Four areas stand out for future quantum computing trends in manufacturing:
- Supply chain optimization
- Materials science
- Product and process innovation
- Predictive maintenance
Supply chain optimization
In manufacturing sectors like aerospace and automotive, supply chains are long and complex, especially when taking into consideration the many tiers that exist. Industry 4.0 technologies are creating a “smart supply chain” awash with data, and quantum computing may be the only way to optimize the movement of materials and products.
Materials science
Conventionally, the process for optimizing existing alloys and developing new ones involves a lot of trial and error, and significant time and expense. Already used for drug discovery, quantum computing can help identify new material compositions that will be stronger and lighter than those available today and do so far faster.
Product and process innovation
Product design, simulation and development, plus process development and inspection, will benefit from the increased computing capabilities offered by quantum. Simulations and analyses will be performed to far higher resolution than is currently practical and there will be more time to evaluate alternatives.
Applied to process optimization problems, the result will be less product variability. Used for quality control in manufacturing, defect prevention and detection capabilities will improve.
Predictive maintenance
Predictive maintenance entails monitoring indicators of machine health for signs of impending failure. A work order initiates maintenance just before a breakdown occurs. The goal is to reduce planned and unplanned downtime by doing just the minimum to keep equipment running at the required rate and quality.
The two main predictive maintenance technologies are machine health monitoring and data analytics. Machine health monitoring entails instrumenting equipment with sensors that measure characteristics like temperature, vibration, and flow. These and other manufacturing sensors, transmit data to computers that store and analyze it to detect trends and sudden changes in condition.
Quantum computing can process far greater volumes of sensor data in less time. This will allow use of more sensors and higher frequency reporting for more granular data on machine operating conditions. In turn, this data will be processed by more sophisticated algorithms, leading to more accurate forecasts. As this builds confidence in the abilities of predictive maintenance, its use will grow and downtime fall.
The role of AI in quantum computing
Artificial Intelligence (AI) involves training and then deploying a model that can identify patterns and trends in data or signals to gain insights not otherwise attainable. AI, especially model training, is computationally intensive, making this an area where quantum computing can enable significant advances in training speed and model sophistication.
AI can also benefit quantum computing by providing models and algorithms that focus processing onto the areas of greatest value. For example, AI can predict and correct errors that arise during quantum computing. AI can also provide the pattern-recognition capabilities missing from quantum computing, which may be particularly useful during intensive image processing and analysis.
Another way AI complements quantum computing is through its ability to make sense of large volumes of data. AI can analyze outputs from quantum computing to provide actionable insights.
Combined with AI, quantum computing can provide valuable insights into processing, scheduling, and optimization problems. It can help with drying times, temperatures and humidities, simulation and by making predictions about equipment and component life.
Future opportunities and challenges
The dramatic increase in computational abilities about to be unleashed by quantum computation promises a transformation in manufacturing. No aspect will be exempt, from product design to process design, scheduling and optimization.
On the product front, new and improved materials will lead to lighter, stronger, longer-lasting, higher-performance products. In parallel, generative design and complex, high resolution simulations will result in products that function better and last longer, while eliminating prototyping and testing activities that slow development.
Entire factories will be recreated as digital twins and used for real-time schedule optimization. Robotic tasks like handling, assembly and packing will be more highly optimized for faster cycle times and reduced error rates. Predictive maintenance will minimize planned downtime and eliminate unplanned repair work.
The step-change in process modeling and optimization, coupled with an explosion in sensor use, will reduce part-to-part variability. At the same time, processes will become faster, more flexible and more efficient creating cost, lead time and customer benefits.
However, such radical changes will not happen overnight and without significant effort. If the manufacturing industry is to take advantage of the potential of quantum computing in advanced manufacturing, education and planning must start today.
Preparing for quantum computing in manufacturing
Quantum computing isn’t readily available to most manufacturers, yet, but the time to start exploring what it could do and how to use it is now. Here are four steps to follow:
1. Educate yourself on the basics. There’s no need to learn quantum mechanics, but learn what quantum computing can do that classical computing can’t and how it can be applied.
2. Develop a plan. Identify, then prioritize, business challenges that need massive computing resources to address. Explore how AI is and will be used and look for ways to integrate it with quantum computing.
3. Investigate quantum computing resources. Organizations such as IBM, Google and D-Wave offer quantum computing services, but using them requires expertise that’s hard to come by. Address this by looking for research partners that can help implement your plan.
4. Invest in R&D. Establish strategic relationships that will help you first explore the potential of quantum computing and then support implementation.
Technology to improve maintenance performance
Manufacturers who want to strengthen their competitive position must stay alert to the potential of new technologies. Of the many that exist, like Industry 4.0 and AI, quantum computing may be the furthest away from practical application, but its impact on manufacturing could be profound. Imagine the benefits of optimizing every activity, from scheduling to predictive maintenance!