Touch Sensing on the Forearm Using the Electrical Impedance Method Yutaro Suzuki * , Kodai Sekimori † , Buntarou Shizuki ‡ , Shin Takahashi § University of Tsukuba, Japan Email: { * ysuzuki, † sekimori}@iplab.cs.tsukuba.ac.jp, { ‡ shizuki, § shin}@cs.tsukuba.ac.jp     Fig. 1: (a) Bands with electrodes attached, to be worn by the user. (b) Touching the forearm. (c) Performing Spider-man. (d) Performing Index pinch. Abstract—We present a novel on-skin touch sensing approach based on the electrical impedance method (EIM). Our approach enables the user to detect touch across the surface of the forearm by wearing two bands, one each on the lower and upper forearm. EIM uses a conductive substance to identify the area being touched. We focused on the electrical conductivity of the skin and applied EIM to the forearm. The two bands have electrodes on the inside. Signals are applied to these electrodes, and the resulting voltage on the surface of the skin is then measured. The advantages of our approach are that it works over a large area of the forearm, and can recognize both hand gestures and touch. Index Terms—Electrical Impedance Method, On-Skin Input, Wearable Devices I. I NTRODUCTION Most wearable devices, such as smart watches, have small interaction areas and thus they have occlusion and fat finger problems [1]. To alleviate these problems, researchers have previously explored the use of human skin as an interactive surface, thus increasing the opportunities for new interactions. Human skin is readily available and quickly accessible, and we can touch our body without looking. Furthermore, compared to wearable devices, skin is sufficiently large for meaningful interaction. Several applications using human skin as an interactive surface have been proposed, such as cameras [2]–[4], infrared sensors [5]–[7], capacitive sensing [8]–[10], acoustic sensing [11]–[13], magnetic sensing [14], [15], and electrical field sensing [16], [17]. However, they often require users to wear an RGB/depth camera, result in occlusion problem, or a limited touch area. In this paper, we propose a touch detection approach that works anywhere on the surface of the forearm. Our system also recognizes hand gestures. To achieve this, we exploited the electrical conductivity of the skin and applied the electrical impedance method (EIM) to the forearm. EIM was originally developed for medical purposes, such as monitoring ventilation and detecting breast cancers. Recently, it has been applied to the field of human-computer interaction (HCI) as an approach for identifying touch positions. In our approach, the user wears two bands. These serve as electrodes on the forearm. The system identifies touches and hand gestures by applying a machine learning algorithm to the voltages detected by the electrodes. Our approach enables touch position identification across a large area of the forearm and has hand gesture recognition capabilities (Fig. 1). II. RELATED WORK In this section, we summarize two related research areas: touch input on the skin, and applications of the EIM. A. On-Skin Input There are many approaches to skin touch sensing. Ogata et al. [2], [3], and Harrison et al. [4] proposed camera- based approaches, in which touch is detected based on images captured by a camera attached to a smart watch or the shoulder. However, they have occlusion problem that user’s hand may hide the touch positions. Skin Buttons [5] and LumiWatch [6] made it possible to de- tect touch around the periphery of a smart watch by attaching infrared sensors to the side. The problem with these approach is that the touch-detectable area is narrow. WatchSense [7] makes it possible to detect touch on the forearm by wearing infrared sensors near the elbow. This approach addresses the problem of the narrow detection area, and the system can detect hover gestures as well as touch. However, occlusion can still be a problem because the actual position touched may be hidden by the user’s hand.