Lucy Turner¹ (), Steven Wiederman², Jessica O'Rielly¹, Anna Ma-Wyatt^3,4; ¹School of Psychology, The University of Adelaide, ²School of Biomedicine, The University of Adelaide, ³Andy Thomas Centre for Space Resources, The University of Adelaide, ⁴CNRS IRL Crossing, The University of Adelaide

The visuomotor system compensates for latencies associated with sensory feedback to enable people to integrate visual information with motor commands and track moving objects with remarkable precision. However, when latency associated with a movement outcome becomes variable or noisy, this ability to compensate is tested. We aimed to identify the thresholds at which these delays disrupt tracking to understand the limits of visuomotor prediction and error correction. In these studies, participants performed a visually guided pursuit task using a mouse, while their eye and cursor movements were recorded under various feedback latency conditions. In the first study, cursor latency was constant across each trial. We systematically tested each of six latency magnitudes (0ms, 100ms, 150ms, 200ms, 250ms and 300ms) and analysed its impact on tracking performance (error and velocity fluctuations). Spectral analysis of cursor velocity showed that pursuit sub movements shifted to lower frequencies interspersed with sporadic high-frequency corrective actions for high-latency conditions. In contrast, smoother, high-frequency tracking dominated under low-latency conditions. In the second study, abrupt changes in cursor latency were introduced mid-trial to examine how sudden variability can influence visuomotor adaptation. Increases in latency resulted in larger positional errors and upregulated corrective sub movements to realign the cursor with the target. Conversely, decreases in latency led to a rapid reduction in positional error and a return to smoother tracking. These rapid transitions in movement highlight the visuomotor system’s ability to recalibrate quickly in response to variability in latency. Together these results show that prediction and error correction mechanisms can differentially respond to changes in latency and this response is dependent on the magnitude of the latency. These results will help inform the design of assistive display technologies that can maintain performance under challenging conditions and increase safety in high-risk environments.