Keio University, a long-standing IBM Quantum Innovation Center in Japan, is advancing quantum reservoir computing (QRC) in collaboration with Mitsubishi Chemical. Their joint research aims to evaluate utility-scale QRC methods by implementing machine learning tasks on IBM quantum processors. This effort builds on a 2023 experiment with collaborators from the University of Tokyo, University of Arizona, and UNSW, where they modeled soft robot motion using quantum input states and linear regression post-processing.
Unlike conventional QRC approaches, which require repeated quantum circuit executions at each time step, the team introduced a repeated measurements technique using auxiliary qubits. This method allows time series data to be extracted in a single execution, significantly reducing runtime and improving prediction accuracy. Experiments were conducted on IBM quantum devices with up to 120 qubits, showing improved performance over traditional QRC schemes and potentially exceeding the capacity of classical simulation.
The Keio-Mitsubishi partnership illustrates how IBM Quantum Innovation Centers support industrial research teams in exploring quantum use cases. Beyond Mitsubishi Chemical, Keio University also works with a broader set of enterprise partners to co-develop quantum algorithms and build in-house quantum expertise. These collaborations enable industry researchers to prototype advanced quantum workflows with guidance from academic experts, accelerating quantum-readiness in practical domains such as materials, finance, and simulation.
Read IBM’s full announcement here.
April 30, 2025
Leave A Comment