Robotic Surgery Input Tracking Correction With Low-Latency LQE

A linear quadratic estimator (LQE) is introduced as an intermediary component between the sensor circuit and the robotic surgery controller. The LQE receives lower-rate, noisier tracking input from the sensor circuit and produces higher-rate, corrected tracking output for the controller. This mediator resolves the contradiction by providing precision through algorithmic correction rather than through hardware sampling rate increases, thereby avoiding the complexity and cost of high-speed sensors.

The LQE dynamically adjusts its internal parameters (prediction coefficients, gain values) based on the characteristics of the tracking input and the desired output sample rate. By changing these parameters adaptively, the system achieves effective noise filtering while maintaining minimal latency. The estimator optimizes the balance between smoothing noisy data and responding quickly to actual movement changes, resolving the contradiction between noise reduction and latency.

The LQE computation is segmented into distinct phases: a prediction phase that runs continuously at the higher output sample rate, and an update phase that executes only when new sensor data arrives. This segmentation allows the system to maintain high output productivity through frequent predictions while reducing computational energy consumption by limiting expensive update calculations to only when necessary. The prediction phase uses lightweight recursive formulas that are computationally efficient.