Supporting Precise Manual-handling Task using Visuo-haptic Interaction

Precise manual handling skills are necessary to create art and to paint models. However, these skills are difficult to learn. Some research has approached this issue using mechanical devices. However, mechanical systems have high costs and limit the user's degree-of-freedom. In our research, we propose a system using visuo-haptics to support accurate work without using any mechanical devices. We considered the principle that when a visuo-haptic force is generated on a user's hand in the opposite direction of a target path, the user moves her/his hand to the right direction reflexively to repel the force. Based on this idea, we created a system that can modify users' hand movement by showing a dummy hand using a mixed reality display, which supports precise manual-handling tasks. To demonstrate this, we performed experiments conducted with a video see-through system that uses a head mounted display (HMD). The results showed that an expansion of the deviation between the target route and the actual hand position improved accuracy up to 50%. We also saw a tendency for a lager expansion to give the most improvement in quality, but slow down working speed at the same time. According to experimental results, we find that a gain of about 2.5 gives an ideal balance between the working precision and the drawing speed.