Insider Brief
- A Japanese research team from Shibaura Institute of Technology, Waseda University, and Fujitsu developed a hybrid quantum–classical method using entangled qubits to accelerate inverse kinematics calculations for robotic arms.
- In simulations, the entangled quantum circuit reached target positions with fewer iterations and higher accuracy than non-entangled quantum circuits and classical methods; testing on a 64-qubit superconducting quantum computer showed a 43% positional error reduction despite hardware noise.
- Published in Scientific Reports, the approach could enable faster, more adaptive robot movements, though current limitations include applicability only to rotational joints, one-way joint influence modeling, and performance impacts from quantum hardware noise.
Entangling quantum bits to represent the joints of a robot arm could make one of robotics’ most time-consuming calculations faster and more precise, according to researchers in Japan.
A team from Shibaura Institute of Technology, Waseda University, and Fujitsu has demonstrated a hybrid quantum–classical approach to solving inverse kinematics — the process of determining which joint angles will put a robot’s arm in a desired position. Inverse kinematics is computationally demanding because robots often have more joints than strictly needed, creating many possible solutions. Classical algorithms must sift through this space, often repeating the process many times during complex motions.
The researchers mapped each robot joint to a quantum bit, or qubit, and introduced entanglement between qubits representing connected links. Entanglement, a property unique to quantum systems, creates correlations between qubits so that measuring one affects the state of the other. In mechanical terms, it mirrors how movement in one link influences another.
In the team’s method, a quantum circuit performs the forward kinematics step, with rotation gates applied to qubits mimicking the motion of real joints. Researchers reported the output is then fed into a classical optimizer, which adjusts the angles until the calculated position matches the target. This cycle repeats until the error falls below a threshold.
Simulations on a two-link arm showed that the entangled quantum circuit reached accurate solutions in fewer iterations than both non-entangled quantum circuits and classical methods. Without entanglement, errors remained around 0.5 meters after 30 iterations. With entanglement, the error dropped sharply after just eight iterations.
The researchers confirmed the approach’s feasibility on a 64-qubit superconducting quantum computer developed by the RIKEN RQC–Fujitsu Collaboration Center. Hardware noise reduced accuracy, but the entangled method still cut positional error by 43% compared with the non-entangled version.
Faster convergence matters in robotics, where systems must often respond in milliseconds to changing conditions. By reducing the number of iterations, the method could allow robots to move more fluidly, avoid obstacles more effectively, and adapt quickly to unexpected forces. While the quantum stage does not replace the classical search entirely, it reshapes the problem space in a way that helps the classical optimizer work more efficiently.
The study, published in Scientific Reports, notes some limitations. The method applies only to rotational joints, not linear ones, and models influence in one direction only — from parent link to child. Hardware noise remains a challenge, and the quantum step adds overhead that must be weighed against the iteration savings.
The researchers propose using the quantum Fourier transform to encode angles more efficiently and parallelizing across links to speed computation. They also note that as quantum devices scale, the technique could be applied to more complex machines, such as humanoid robots.
“By leveraging quantum circuits for forward and inverse kinematics, our approach offers a promising path toward scalable and efficient robotic computation, with further enhancement expected through advanced entanglement-based quantum algorithms,” the team noted.
