URSI GASS 2021, Rome, Italy, 28 August - 4 September 2021 Multi-Channel Radiofrequency Finger Augmentation Device for Tactile Internet Federica Naccarata (1) , Giulio M. Bianco* (1) , and Gaetano Marrocco (1) (1) Pervasive Electromagnetics Lab, University of Rome Tor Vergata, Rome, Italy Abstract The Tactile Internet (TI) aims to enhance human-to- machine interactions by transmitting the human senses through the internet. Promising candidates to enable TI applications in the short term are the Radiofrequency Finger Augmentation Devices (R-FADs), hand-worn epidermal RFID (Radiofrequency Identification) systems. R-FADs were employed to recover the thermal feeling and discriminate dielectric materials. Preliminary experiments only considered single-channel R-FADs, whereas multi- channel systems with multiple tags could more reliably sense and communicate with the reader. In this work, a multi-tag R-FAD is for the first time manufactured and tested. The five tags’ combined response, named digital fingerprint, allows for more reliable discrimination of the tested materials. 1 Introduction The Tactile Internet (TI) is an emerging technology that aims to revolutionize human-to-machine interaction [1] by spreading the human senses through the Internet. TI is expected to be boosted by 5G mobile communication networks [2] which can handle massive data traffic with extremely low end-to-end latencies so that the user is not affected by cybersickness [3]. An interesting application is the recovery of damaged sense, but the implementation of new kinds of perception denoted as sensorial ultrability [4], is possible, too. The enabling technology for TI is that of Finger Augmentation Devices (FADs) [4] based on Epidermal Electronics. Moreover, waiting for 5G technology to become more mature, UHF Radiofrequency Identification (RFID) can be exploited in the short term to establish communication links through backscattering modulation, avoiding batteries on the tags. Using Epidermal Electronics, soft and stretchable devices can be fabricated for application on the fingertips so that the user’s natural gestures are not hindered. Recently, Radiofrequency FADs [5] (hereafter denoted as R-FADs) were proposed for the restoration of peripheral thermal feeling in impaired people [6]. R-FADs were then empowered by including a new family of ICs (Integrated Circuits) having auto-tune capability. Namely, they can automatically modify their internal impedance to make the matching to the hosting tag antenna nearly insensitive to changes of local boundary conditions. A constrained method for designing these devices was Figure 1. Concept of the multi-channel R-FAD system comprising stretchable epidermal tags over the fingertips and an interrogating antenna on the back-hand. proposed in [7] and then applied to recognize some liquids filling a bottle, with interesting applications to aid blind people. In all cases, just a single fingertip was sensorized and, hence, the real potentiality of this framework is still mostly under-used. This paper explores, for the first time, a multi- channel R-FAD, wherein all the five fingertips are simultaneously sensorized, so that richer information is available to recognize the touched objects. Having recalled the rationale for dielectric sensing with auto-tuning ICs, an experimental setup comprising stretchable fingertip antennas is applied in a test campaign involving volunteers touching different materials. The goal is to evaluate the inter-user variability of the collected data, and, most importantly, identify the digital contrast of the measurements with respect to a single-finger sensing. 2 Rationale of Dielectric Sensing by Auto-tuning chips The core of an auto-tuning RFID IC [7] can be modelled as an adaptive internal network of parallel capacitors, whose total capacitance depends on the number ݏof activated components: ܥ (ݏ)= ܥ ௠௜௡ + ܥݏ (1) where ܥ ௠௜௡ and ܥ are specific parameters of the IC. The resulting index saturates outside the range ௠௜௡ ݏ ௠௔௫ , where ௠௜௡ , and ௠௔௫ depends on the implementation of the IC. The equivalent susceptance