By Carolyn Mathas

IonQ and Hyundai just announced an expansion of their partnership. They will develop machine vision algorithms capable of conducting object detection on 3-D data from autonomous vehicles and simulate electrochemical reactions of a variety of metal catalysts.

According to Sonika Johri, Lead Quantum Applications Research Scientist at IonQ, “With Hyundai, we are looking at machine learning and chemistry applications. In both areas, many quantum algorithms have been proposed and analyzed theoretically, but there are few demonstrations that go beyond a handful of qubits and gates.” Johri explained that, by using #AQ 25, “…we can study these algorithms on real hardware, doing analysis of how the accuracy and time-to-solution of the algorithms will scale in real life.” Another benefit is it allows the companies to iterate between co-designing the algorithms and hardware to reach real-life performance beyond theoretical analysis.

Two challenges that exist involve chemistry and object detection. “The development of next gen batteries for electric vehicles (EVs) requires engineers to have a deep understanding of the battery chemical reaction mechanisms so that we can identify undesired reaction paths that hurts the battery efficiency or find efficient and cheaper catalysts to improve battery efficiency,” Johri said. “Simulating complex chemical systems to high accuracy is extremely difficult on classical computers.” IonQ is working with Hyundai to develop quantum algorithms capable of efficiently simulating the battery chemical systems on IonQ’s ion-trapped quantum computers. 

For an autonomous vehicle, accurately in identifying the location and category of an object in its surroundings is critical. “Data from a variety of sensors such as cameras or LiDAR may be used for this purpose. IonQ is working with Hyundai on the diverse datasets to improve the reliability, accuracy and efficiency of the models.”

The partnership expansion is based on *an entirely new approach* to approaching problems using quantum computers. For example, when digital computers were first capable of crunching numbers, theoretical chemists trying to understand chemical reactions had no idea how to use the capability. A new computational chemistry framework was developed taking advantage of number crunching machines not previously available, resulting in a new field of computational chemistry with more powerful insights and understanding of chemical reactions.

“In a sense, one can consider our current efforts as the process of developing new approach to understanding chemistry or machine learning in ways that we are not able to do today. It has the potential to transform the way we tackle these hard problems that are not practical using traditional computational methods,” said Johri.

Quantum machine learning techniques at IonQ are yielding the potential to learn faster, be more effective in recognizing edge cases, generalize better, learn using lower resolution or noisy data, and capture complex correlations using a lower number of parameters.

Hyundai’s earlier efforts to study lithium compounds and the chemical reactions involved in battery chemistry, provided a way to explore new metal catalyst chemical reactions for future vehicles. Collective insights and knowledge gained from quantum simulations will enable Hyundai engineers to develop higher-performance EVs at lower costs.

Additional information about this agreement can be found in a news release posted on the IonQ website here.

December 7, 2022