Electrical muscle stimulation (EMS) became a popular method for force-feedback without mechanical-actuators. While much has been written about the advantages of EMS, not much work has investigated circumventing its key limitations: (1) as impulses traverse the skin, they cause an uncomfortable “tingling”; (2) impulses are delivered via gelled-electrodes, which not only require direct skin contact (must be worn under clothes); but, also (3) dry up after a few hours. To tackle these, we explore switching from electrical to magnetic muscle stimulation (MMS), via electromagnetic fields generated by coils. The first advantage is that MMS coils do not require direct skin contact and can actuate up to 5 cm away (Study#1)—this enables applications not possible with EMS, such as stimulation over the clothes and without ever replacing electrodes. Second, and more important, MMS results in ∼50 % less discomfort caused by tingling than EMS (Study#2). We found that reducing this tingling discomfort has two downstream effects for interactive systems: (1) participants rated MMS force-feedback as more realistic than that of EMS (Study#3); and (2) participants could more accurately perceive the pose actuated by the interactive system (Study#4). Finally, we demonstrated applications where our proposed switch from EMS to MMS improves user experience, including for VR feedback, gaming, and pose-control.

Smartwatches gained popularity in the mainstream, making them into today’s de-facto wearables. Despite advancements in sensing, haptics on smartwatches is still restricted to tactile feedback (e.g., vibration). Most smartwatch-sized actuators cannot render strong force-feedback. Simultaneously, electrical muscle stimulation (EMS) promises compact force-feedback but, to actuate fingers requires users to wear many electrodes on their forearms. While forearm electrodes provide good accuracy, they detract EMS from being a practical force-feedback interface. To address this, we propose moving the electrodes to the wrist—conveniently packing them in the backside of a smartwatch. In our first study, we found that by cross-sectionally stimulating the wrist in 1,728 trials, we can actuate thumb extension, index extension & flexion, middle flexion, pinky flexion, and wrist flexion. Following, we engineered a compact EMS that integrates directly into a smartwatch’s wristband (with a custom stimulator, electrodes, demultiplexers, and communication). In our second study, we found that participants could calibrate our device by themselves Math 1 faster than with conventional EMS. Furthermore, all participants preferred the experience of this device, especially for its social acceptability & practicality. We believe that our approach opens new applications for smartwatch-based interactions, such as haptic assistance during everyday tasks.

Figure 1: Left: positions of vibrators and trackers