Steering is initiated based on error accumulation

Abstract

Vehicle control by humans is possible because the central nervous system is capable of using visual information to produce complex sensorimotor actions. Drivers must monitor errors and initiate steering corrections of appropriate magnitude and timing to maintain a safe lane position. The perceptual mechanisms determining how a driver processes visual information and initiates steering corrections remain unclear. Previous research suggests two potential alternative mechanisms for responding to errors: (i) perceptual evidence (error) satisficing fixed constant thresholds (Threshold), or (ii) the integration of perceptual evidence over time (Accumulator). To distinguish between these mechanisms an experiment was conducted using a computer-generated steering correction paradigm. Drivers (N=20) steered towards an intermittently appearing ‘road-line’ that varied in position and orientation with respect to the driver’s position and trajectory. One key prediction from a Threshold framework is a fixed absolute error response across conditions regardless of the rate of error development, whereas the Accumulator framework predicts that drivers would respond to larger absolute errors when the error signal develops at a faster rate. Results were consistent with an Accumulator framework, thus we propose that models of steering should integrate perceived control error over time in order to accurately capture human perceptual performance.

Error accumulation when steering toward curves

“…This allowed the paradigm to directly exploit the key theoretical differences between Threshold and Accumulator frameworks: how a perceptual signal builds over time. Goodridge et al (2022) found that the timing and magnitude of steering behaviours were in line with Accumulator predicted responses. Drivers did not respond based 7 upon time-independent thresholds, rather they altered their response to the rate at which the perceived control error developed.…”

“…Two alternative frameworks -Threshold and Accumulatorhave been proposed as mechanisms that could describe the intermittency involved in steering. Whilst previous research has demonstrated that steering toward straight road-line targets is best explained via an Accumulator framework (Goodridge et al, 2022), aspects of the experimental design (discussed in more detail later) may have made it more likely that people would accumulate perceptual information rather than rely upon fixed time independent thresholds. Therefore a specific aim of this manuscript was to build upon the work conducted by Goodridge et al (2022) to investigate whether Accumulator-predicted steering responses translate to a more general context where the optical information presented to drivers is more closely aligned to that experienced during real-world locomotor settings (i.e.…”

“…Hence a driver may only produce a braking response once optical expansion surpasses some fixed magnitude. Whilst it may seem intuitive that a human would initiate a sensorimotor action based directly upon the perceptual information that is presented at a particular instant, recent evidence within the driving domain has suggested that this is too simplistic to reliably replicate human performance (Durrani et al, 2021;Goodridge et al, 2022;Markkula et al, 2018Markkula et al, , 2021. When braking in response to a looming signal, it has been found that drivers do not simply respond once the looming surpasses a fixed value.…”

“…In Goodridge et al (2022) the locomotor conditions initially simulated a linear direction of travel relative to a visible straight road-line that could be offset at one of a number of possible orientations. However, such a set up produced a prominent egocentric visual angle 𝛼 when the line was first presented when positioned on the road-line (during 0 m starting position conditions, see Figure 2A).…”

“…Hence previous literature has suggested that altering an observers perceived heading can affect the strategy observer's use to steer, and the precision with which they are able to implement their locomotor strategy. Furthermore, counter-rotating the virtual heading in the manner produced by Goodridge et al (2022) could have produced a sensation of vehicle drift. This is because the optical information participants received (specified by the camera view) had an angular offset relative to the direction they were travelling.…”