Takashige Suzuki, Kizashi Nakano, Takuji Narumi, Hideaki Kuzuoka
Virtual Reality (VR) training systems often rely on visual cues, which can compete for the user’s attention, particularly in high-skill domains like surgery or assembly. While vibrotactile guidance offers a non-visual alternative, current single-limb systems suffer from limited spatial resolution, restricting their directional precision for complex tasks. To overcome this limitation, we propose a multi-limb “Tacton (symbolic vibrotactile patterns)” strategy that distributes vibrotactile information across anatomically distinct limbs (the wrist and ankle). We conducted two experiments to validate this approach and determine the optimal reference frame for inter-limb coordination. Experiment 1 ( N = 12 ) evaluated the distribution strategy by comparing the directional precision of single-limb (wrist and forearm) versus multi-limb (wrist and ankle) configurations using a novel temporal pattern encoding for 32 unique directions. Results demonstrated that distributing cues significantly improved precision, reducing angular errors to under 22.5 ° in over 77% of trials compared to the single-limb condition. Experiment 2 ( N = 12 ) addressed the cognitive challenge of coordinating these distributed signals by comparing a body-based “Skeletal” frame with an environment-based “World” reference frame. The “World” frame, which maps cues to an allocentric coordinate system, yielded substantially faster reaction times and lower angular errors than the “Skeletal” frame, minimizing the cognitive load associated with mental rotation. We conclude that high-precision, non-visual hand guidance is best achieved by distributing symbolic haptic cues across separate limbs and mapping them to a stable, allocentric coordinate system. These findings provide foundational design principles for creating immersive, hands-free guidance systems that preserve the user’s visual-attentional resources.