Can Tactile Information be Used as a Cue in Time-to-Collision Estimation?

Abstract

We measured time-to-collision (TTC) judgments of (imaginary) approaching objects based on a 250-Hz vibrotactile stimulus delivered to the fingertips by a C-2 tactor. Changes in stimulus intensity simulated a sound source that approached a receiver. The vibration stopped after 1 s when the (imaginary) object’s actual TTC was between 0.5 s and 6 s. Participants were told to press a key when they thought the object would reach them. They also were told that intensity signified the object’s distance (less intense is farther). Mean estimated TTC increased as actual TTC increased, as it does in auditory and visual modalities, but the relationship was more compressed with tactile stimuli. Judgments were driven by final vibration intensity, rather than the object’s presented TTC, as we reported previously with auditory TTC estimation. Results have implications for the design of technologies used to aid drivers or pedestrians avoid collisions, particularly those with vision loss.

The ability to avoid collisions is critical for activities such as driving a car or walking across a road. To avoid collisions, people must discern that a collision might occur and estimate how much time remains before the collision would occur (time-to-collision; TTC) so that they can respond at the right time to avoid it. Earlier investigations of collision perception showed that people can use visual (Lee, 1976, 1980) and auditory information (Shaw et al., 1991) in TTC estimations of approaching objects but did not examine the potential usefulness of tactile information.

Haptic devices have been developed to help visually impaired people avoid collisions by presenting distance information through vibration (e.g., Miniguide Mobility Aid: Hill & Black, 2003; Sunu Band: Tang & Marinoff, 2022), presuming that people can learn an association between an object’s distance and the intensity or rate of vibration. However, it is not known whether people use tactile information for collision avoidance with such devices or, in particular, whether vibrotactile signals can be used to judge the temporal distance to an object (i.e., TTC). We examined whether people could use vibrotactile information to estimate TTC and whether results are similar to TTC estimates in vision and audition.

We measured time-to-collision (TTC) judgments of (imaginary) approaching objects based on a 250-Hz sinusoid vibrotactile stimulus presented to the fingertips by a C-2 tactor. Changes in stimulus intensity simulated a sound source that approached the observer at a constant speed. The vibration stopped after 1 s when the (imaginary) object’s actual TTC was 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, or 6.0 s. Twenty-one undergraduates from Rice University pressed a key when they thought the object would reach them. They were told that intensity signified the object’s distance; more intense vibrations represented closer distances than less intense vibrations. Within each level of TTC, the object started “near” and approached relatively slow or started “far” and approached relatively fast. This combination of TTC and final distance resulted in 14 unique conditions; each was presented 10 times in a random order. TTC judgments were measured as the time interval between the end of the tactile stimulus and the participant’s key press.

Results of a two-way repeated-measures ANOVA indicated effects of TTC, final distance and their interaction, p < .001. Mean estimated TTC increased as actual TTC increased, as it does in auditory and visual modalities (DeLucia et al., 2016; Keshavarz et al., 2017). However, the relationship was much more compressed in the tactile modality, even though the intensity profiles were identical to the auditory stimuli from our previous TTC estimation study (DeLucia et al., 2016). In fact, average estimated TTC remained virtually the same for actual (imaginary) TTCs between 3.0 s and 6.0 s.

At each actual TTC, significantly longer TTCs were estimated for the “far”/”fast” objects compared to the “near”/”slow” objects. This can be attributed to the fact that far objects had a lower final vibration intensity than near objects. This “intensity-arrival effect” has been reported for auditory TTC estimation (DeLucia et al., 2016; Oberfeld et al., 2022). Inspection of the data showed that the average estimated TTCs were almost perfectly consistent with a TTC estimation strategy based solely on the final intensity of the vibrotactile stimulus, as we found in our prior study on auditory TTC estimation (DeLucia et al., 2016). Results have implications for the design of technologies used to assist drivers or pedestrians avoid collisions, especially those with vision loss. Mobility aids should consider the compressed relationship between estimated and actual TTC that occurs after 2.5 s. One limitation of the current study is that only vibration intensity was used to simulate approach. Future studies should examine the effectiveness of vibration frequency.