Quantum computing achieves superiority in the automotive industry Quantum artificial vision is used for the first time in parts inspection

Why does the automotive industry need quantum computing? The automotive industry is in the midst of a once-in-a-century transformation, with four major trends that will transform the global automotive industry, namely "CASE" - Connected, Autonomous, Share & Service, and Electric. Share & Service), and Electrification (Electric). According to McKinsey's report, quantum computing will permeate all aspects of these automotive industries.

Quantum Computing May Enhance Key Steps in the Entire Automotive Production Automation Value Chain

A joint study by Multiverse Computing and IKERLAN recently showed that image classification by quantum artificial vision systems can detect defects in automotive part manufacturing with better results than common classical methods. a preprint version of the paper was submitted on August 9 on the arXiv website under the title "Quantum artificial vision for manufacturing defect detection" [1].

Specifically, the research team developed a quantum-enhanced kernel method (kernel method) for classification on IBM's general-purpose gate-based quantum computer, the Quantum Support Vector Machine (QSVM), and a quantum classification algorithm on the D-Wave quantum annealer --The researchers found that both algorithms outperformed common classical methods in identifying relevant images and accurately classifying manufacturing defects.

Examples of images of the "Castings" group to be identified in the experiments, and the corresponding borders surrounding the defects. All experiments include 2727 images of automotive parts with/without casting defects.

Graph of experimental results. The results presented here are benchmark tests of the algorithms for quantum and classical image classification in three cases: (i) when including/excluding dimensionality reduction via PCA, (ii) when enhancing the contrast of the image, and (iii) when changing the scale in the dataset. Left table: running time without dimensionality reduction processing; middle table: results after applying contrast stretching to the images; right table: results after adaptive equalization processing to the images. The fastest classification method is bolded in bold.

The QBoost algorithm gives impressive results, which highlights the capability of current quantum annealing techniques. More importantly, while QSVM requires a quantum computer to perform the training and inference tasks, the QBoost-based solution is able to train the model in a quantum system and then deploy the model in a classical system. This makes it possible in the near future to utilize quantum-based methods in real industrial manufacturing scenarios without the need to adapt the ecosystem to deploy and use them.

Commenting on the results, Ion Etxeberria, CEO of IKERLAN, said [2], "To the best of our knowledge, this research represents the first implementation of a problem related to quantum computer vision in manufacturing production lines. This collaborative research confirms the benefits of applying quantum methods to real-world industrial challenges. We strongly believe that quantum computing will play a key role in providing artificial intelligence-based solutions for particularly complex scenarios."

"Quantum machine learning will dramatically disrupt the automotive and manufacturing industries," said Dr. Roman Orus, chief scientific officer of Multiverse Computing, "We are excited to see the value of early applications of quantum computing today, such as quantum artificial vision, and are pleased enter a new era of machine learning with forward-thinking partners such as IKERLAN as quantum technology continues to evolve."

About IKERLAN.

IKERLAN was founded in 1974 to provide competitive value to the industry. It offers holistic solutions that integrate two major, disparate technology areas: electronics, information and communication technology (EICT) and energy and mechatronics. The organization is a partner member of MONDRAGON and the Basque Research and Technology Alliance (BRTA).

About Multiverse Computing.

Multiverse Computing is a leading quantum software company that applies quantum and quantum-inspired solutions to solve complex problems in finance to create value today and enable a more resilient and prosperous economy. Its flagship product, Singularity, allows professionals across a wide range of industries to leverage quantum computing through common software tools. The company also serves companies in the mobile, energy, life sciences and Industry 4.0 sectors.

Reference links：

[1]https://arxiv.org/abs/2208.04988

[2]https://multiversecomputing.com/resources/multiverse-computing-and-ikerlan-detect-defects-in-manufacturing-with-quantum-computing-vision