Autopilot coming? Hyundai will use quantum computing to speed up the process

On April 19, IonQ and Hyundai Motor Corporation announced a new project to use quantum machine learning to improve computational processes for tasks such as classification and simulation of road sign images in real-world test environments, extending to 3D object detection.

This is the second project involving quantum computing that IonQ has announced with Hyundai Motor in 2022. In January 2022, they announced a study to explore the use of quantum computing to improve battery chemistry.

It's no secret that as car companies move to self-driving, semi-autonomous vehicles, computational complexity needs to increase dramatically. While it is unlikely that anyone will install a full-fledged quantum computer in a car in recent decades, quantum computers still hold enormous potential for accelerating the training of machine learning models for various automotive-related machine learning applications. Training a machine learning model is the most computationally intensive job in machine learning. It consists of large datasets that provide models with underlying situations and uses sophisticated algorithms to create mathematical models that can be used to identify features in images.

Image classification and 3D object detection are fundamental steps towards the next generation of mobility, including autonomous vehicles. Together, IonQ and Hyundai will seek to boost computing power through more efficient machine learning on quantum computers, as they can process large amounts of data faster and more accurately than classical systems.

Example of a neural network that can distinguish cats from dogs in pictures

Leveraging a breakthrough in encoding images into quantum states, IonQ is already using IonQ's quantum processor to classify 43 types of road signs. In the next phase, the two companies will apply IonQ's machine learning data to test environments in modern vehicles and simulate various real-world scenarios.

In the latest project, IonQ and Hyundai will use their Aria quantum computer to help create a machine learning model for classifying 43 types of road signs commonly found in mobile situations. While the training of the model does not take place in the car, once created with the model, it can be performed using an on-board classical computer that receives input from the on-board camera and determines if any road signs are present in the image.

The initial phase of the program will be to use IonQ's "Aria processor" - the IonQ Aria, with 20 algorithmic qubits (#AQ), is the industry's most powerful quantum computer based on application-oriented standard industry benchmarks. Create models with Aria, then bring them to a modern test environment and simulate a variety of real-world scenarios. Longer-term activities could include expanding research to include more general 3D object detection so that the system can identify pedestrians or cyclists. For this project, IonQ will use a breakthrough algorithm to encode images into quantum states.

“We are excited to expand our existing collaboration with Hyundai to focus on another key aspect of next-generation mobility,” said Peter Chapman, President and CEO of IonQ, “From collaborations on electric vehicle battery research to image classification for autonomous driving and object detection research, we hope to see quantum computers become an integral part of developing new transportation solutions."

Link:

[1]https://quantumcomputingreport.com/ionq-collaborates-with-hyundai-on-a-quantum-machine-learning-application-for-image-classification-and-3d-object-detection/

[2] https://ionq.com/news/april-19-2022-ionq-hyundai-quantum-machine-learning