Proceedings of The Physiological Society
University of Oxford (2011) Proc Physiol Soc 23, PC88
Do physiological time delays provide the reason for intermittent, visual-manual control?
C. van de Kamp1, I. D. Loram1
1. IRM, Manchester Metropolitan University, Manchester, United Kingdom.
It has been almost exclusively assumed that a continuous paradigm from engineering control theory is most suitable in explaining sustained control of human movements. However, recently, it has been shown that controlling an unstable load serially ballistically through intermittent taps applied to a joystick is natural and more effective than using continuous hand contact (Loram et al., 2011). Furthermore, it was found that, when maximizing positional control, participants adopted a Median Tapping Interval (MTI) of 0.450 (0.2 iqr) s which corresponds to two taps/s. However, the physiological basis of this interval is unknown. Here we studied whether the MTI is determined by intrinsic physiological factors (such as the feedback time delay) or by external load dependent factors such as the complexity (order) or stability of the load being controlled. In ten different conditions ten participants, using a joystick, controlled the position of a dot (representing the position of a virtual load running in real time) on an oscilloscope for 200s. Participants were asked to keep the load position close to the centre of the screen by keeping either continuous or intermittent contact with the joystick. In eight trials the joystick position modulated the acceleration of the load (2nd order) whereas in the remaining two trials it specified the velocity of the load (1st order). Within the 2nd order load conditions we varied the stability of the load by changing what might be thought of as the passive stabilization (85% vs 20%) of the gravitational moment acting on an inverted pendulum. In the six 2nd order 85% stabilized load conditions the transfer of joystick position (control force applied to the load) was delayed for: 0, 125, or 250 ms. MTI were calculated from the joystick movements and compared by ANOVAs. MTI were found to increase significantly with increasing levels of artificially added delay. When there was no added delay, the median duration between taps was 0.375 (0.13 iqr)s which corresponds to two-three taps per second. When we added a delay 125 ms we found a MTI of 0.450 (0.18)s (two taps/s) and when participants had to deal with delays of 250 ms the MTI increased up to 0.550 (0.28)s (< two taps/s). No effects of load stability or load order were found. Increasing the time delays resulted in slower tapping rates. Since we can exclude other factors such as load complexity (order) and unstable time constants (stability) we conclude that the (intrinsic) time delay is the physiological factor determining the best intermittent interval. In engineering, intermittent controllers (cf. Gawthrop et al., 2011)) were designed to deal with (long) time delays which are problematic for a continuous control model. These intermittent control models for machines provide a new physiological paradigm for understanding sustained control of posture and movement in man.
Where applicable, experiments conform with Society ethical requirements