Researchers in Spain have used two different quantum computing techniques for machine learning in industrial vision systems.
The research into quantum artificial vision for defect detection in manufacturing used X ray images from an automotive production line. These were compared to existing machine learning image classification algorithms.
The researchers from Multiverse Computing and Ikerlan in Spain developed a quantum-enhanced kernel method for classification on universal gate-based quantum computers as well as a quantum classification algorithm on a quantum annealer. They found that both algorithms outperformed common classical methods in the identification of relevant images and the accurate classification of manufacturing defects.
The quantum computer analysis simulated IBM’s 16 qbit machine with a quantum support vector machine (QSVM) algorithm while the Qboost quantum annealing algorithm was run on D-Wave’s quantum system. Both showed significantly faster operation than the classical systems.
- Quantum deal to look at semiconductor materials
- Bosch to build quantum digital twin of Madrid factory
- IBM aims for 4000 qubit quantum computer
The quantum vision systems are benchmarked for an unbalanced dataset of images, with QBoost allowing for larger problems to be analyzed with present-day quantum annealers. The team looked at data preprocessing, including dimensionality reduction and contrast enhancement, as well as hyperparameter tuning in QBoost.
“To the best of our knowledge, this research represents the first implementation of quantum computer vision for a relevant problem in a manufacturing production line,” said Ion Etxeberria, CEO of Ikerlan, a technology transfer centre consultancy. “This collaborative study confirmed the benefits of applying quantum methods to real-world industrial challenges. We strongly believe that quantum computing will play a key role in providing AI-based solutions to particularly complex scenarios.”
“Quantum machine learning will significantly disrupt the automotive and manufacturing industries,” said Roman Orus, Chief Scientific Officer at Spanish quantum software firm Multiverse Computing and Visiting Professor at the Barcelona Supercomputing Centre in Barcelona, Spain. “We are pleased to witness the value of early applications quantum computing today, such as quantum artificial vision, and excited to enter a new era of machine learning.”
The research paper is here: arxiv.org/abs/2208.04988
Related quantum computing articles
- D-Wave steps up battle for quantum computers
- CEO Interview: Building a quantum computing leader in Europe
- Quantinuum extends UK quantum UK hub with IBM deal