Low-Cost Millimeter-Wave Interactive Sensing through Origami Reflectors Riku Arakawa Yang Zhang The University of Tokyo University of California, Los Angeles riku.arakawa1996@gmail.com yang.zhang@cs.cmu.edu mm-wave sensor order controller kit fold origami calibrate controllers use controllers Figure 1: We envision a future scenario where users can (a) buy cheap materials (e.g., origami), (b) assemble to make controllers by themselves, (c) calibrate the controllers in their own environment guided by a dedicated app, and (d) deploy them as low-cost controllers for achieving ubiquitous interactivity. Please also see our Video Figure1 for more details. ABSTRACT 1 INTRODUCTION Millimeter-Wave (mm-wave) sensing provides an increasingly vi- Millimeter-Wave (mm-wave) sensing poses an inviting opportunity able sensing solution for smart environments for its compact and for ubiquitous sensing for being low-cost, compact, and privacy- solid-state form factor, non-intrusiveness, and low cost. While prior sensitive – key properties that make commercial integrations possi- work in this domain has mostly focused on sensing humans – e.g., ble. Recent years have seen an increasing trend of mm-wave sensing location, motion, and posture, we propose a new approach that techniques featured on consumer products such as smartwatches, leverage mm-wave sensing to enable tangible ubiquitous controllers phones, autonomous vehicles, as well as home devices such as such as buttons and switches. By encoding the controller state with occupancy sensors, smart lightbulbs, and thermometers. In these the Radar Cross Section (RCS) of origami structures, our component- systems, mm-wave sensors emit structured RF waves into user free controllers cost less than 40 cents per unit and require virtually environments and decode the reflectance signals to infer user infor- zero maintenance effort, while achieving long-range wireless sens- mation such as presence, proximity, hand gestures, body postures, ing with sufficient accuracies. and beyond. Meanwhile, conventional sensing techniques for ubiquitous in- teractivity have several constraints that have prevented smart en- CCS CONCEPTS vironments from being widely adopted across society. Existing • Human-centered computing → Ubiquitous and mobile com- controllers such as switches and buttons are often wired, elimi- puting systems and tools. nating flexible deployments. Though, it is possible to enable flex- ible deployments with batteries and wireless transceivers, these components inevitably increase the maintenance and monetary KEYWORDS costs. In response, prior work proposed interactive sensing tech- Millimeter Wave, Ubiquitous Computing, Battery-free, DIY niques around computer vision [27], capacitive sensing [32], and RF backscatter [14]. Researchers have also leveraged user interactions as sources of power to eliminate the need for batteries [3, 24]. In this work, we leverage the increasingly popular mm-wave sensing technique to build wireless interactive controllers that are designed around origami-inspired structures. While most prior work on mm-wave sensing focused on sensing direct signals from Copyright 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). users, our system senses objects – controllers that encode user interactions into their radar cross section (RCS), which we can detect to infer user interactions wirelessly at a room-scale. More- over, our controllers consist of only common everyday materials 1 https://youtu.be/BaLlQNVRVrI CHIIoT 1, February 17, 2021, Delft, The Netherlands Riku Arakawa and Yang Zhang (e.g., cardboard papers, aluminum foil), which are low-cost and several commercial products such as QIACHIP switches [21] and component-free, enabling users to deploy them across their envi- EnOcean switches [4]. ronments with negligible cost or maintenance effort. In this paper, we propose an approach for achieving battery-free Specifically, we designed a series of origami-inspired controllers wireless controllers that consist of only ultra-low-cost everyday to implement conventional controls, including ones with discrete materials in origami forms, working in conjunction with mm-wave states such as buttons and switches, as well as ones with continuous sensing. The proposed approach enables users to deploy controllers states such as knobs and sliders. We built a signal processing and de- with negligible cost and effort to maintain. Furthermore, in the tection pipeline around a Frequency-Modulated Continuous Wave future, we expect our work to facilitate users making and replicating (FMCW) radar and conducted a series of evaluations to demonstrate these paper-based controllers easily on their own, enabling DIY our system’s feasibility. Overall, we believe our system proposes a smart environment experience. promising approach to achieving ubiquitous interactivity. 3.2 RF Sensing in HCI 2 INSPIRATIONS On the technology front, our approach is closely related to systems that leverage RF sensing. RF sensing has been actively integrated This work took much inspiration from prior literature, including into applications in our lives. One of its representative uses is recent effort in HCI that leveraged paper as interaction medium, in autonomous vehicles where mm-wave radar is used primarily where researchers empowered paper with sensing [14, 30] and for detecting objects around the vehicle (e.g., other vehicles and power generation [2, 11]. We were also inspired by a vast amount pedestrians) [8, 26, 35]. Recently, Prabhakara et al. [20] showed of online resources on origami structures in the design of our con- that a wireless approach to sensing tire wearing is also possible by trollers. Finally, our project chose to follow the same low-cost and using mm-wave. do-it-yourself spirit as Nintendo Labo [17], a phenomenal inter- Closer to our work are prior systems that focus on interactivity action design concept for console games. For these reasons, we in HCI. For example, Soli [15] is a ubiquitous gesture sensing tech- chose paper, one of the most accessible and tangible materials to nology based on FMCW mm-wave sensing, which allows precise build our controllers. We envision a future where these paper-based finger tracking at close range [25]. The same mm-wave sensing controllers can be easily accessed and assembled by average home- has also been applied for classifying proximate material and object owners to facilitate smart environment interactivity (Figure 1). [28, 29]. Moreover, similar FMCW-based sensing using RF signals are proposed for capturing human pose and motion even when they 3 RELATED WORK are occluded from the device or in a different room [1, 33, 34]. These To situate our work, we first review previous approaches proposed approaches have been expanded to be capable of identifying users for room-scale interactions. We also review how RF sensing has by analyzing the signal reflections and been utilized for collecting been utilized to power interactions in HCI. behavioral data in homes [6]. Ultra-wideband (UWB) radar has been another common RF sens- ing technique used for various purposes such as localizing sur- 3.1 Room-Scale Interactions rounding IoT devices and enhancing interactions with them [7, 10]. Many previous works have aimed to achieve room-scale interac- MechanoBeat [18] is a low-cost mechanical tag that can work with tivity by deploying ubiquitous sensing modules based on various UWB radar arrays for unobtrusively monitoring user interactions. principles. Vision-based sensing approaches have been widely pro- Additionally, the RF Doppler effect can also be utilized for detecting posed [13, 27]. For example, WorldKit [27] is a system that uses movements of targets. For example, Goel et al. [5] leveraged the a depth camera and a projector to make ordinary surfaces (e.g., effect to detect facial gestures by monitoring user’s tongue, cheeks, walls) interactive. SurfaceSight [13] is a LiDAR-based sensing sys- and jaw movements. tem that enriches IoT experiences by enabling sensing context on As shown in these works, while most previous works have been table surfaces. On the other hand, sensing approaches without re- focusing on sensing direct signals from users, we aim to sense the lying on vision sensors have also been investigated. Wall++ [32] states of objects as controllers that encode user interaction. In detail, is a capacitive sensing approach for allowing walls to become a we propose an approach to utilizing origami-inspired structures as smart infrastructure that senses users’ touch and gestures. More- ultra-low-cost reflectors that can change their shapes upon user over, lasers have been utilized for sensing room-scale interactions interaction. We expect that this shape change can result in unique from a distance [19, 31]. RCS values which can be sensed remotely with a mm-wave radar. Meanwhile, there are also wireless sensors which allow flex- In Section 5, we explain how our origami-based controllers were ible installation and portable use (e.g., TV remotes). Still, these designed. controllers are often battery-powered, which requires user main- tenance (e.g., battery replacement). In response, there have been 4 SENSING HARDWARE developed battery-free wireless controllers. For example, The Pep- We used the Infineon Position2Go 2 , a 24GHz radar sensor devel- permill [24] utilizes human operation as a source of power, and opment kit utilizing BGT24MTR12 RF transceiver and XMC4700 PaperID [14] uses RF backscatter to sense how a user is manipulat- 32-bit ARM® Cortex®-M4 MCU series, which costs approximately ing RFID-instrumented paper. In addition, Iyer et al. [9] embedded $300. The sensor has one transmitter (Tx) and two receivers (Rxs) backscatter structures into 3d-printed objects to make wireless sen- sors such as buttons, knobs, and sliders. Finally, there have been 2 https://www.infineon.com/cms/en/product/evaluation-boards/demo-position2go/ Low-Cost Millimeter-Wave Interactive Sensing through Origami Reflectors CHIIoT 1, February 17, 2021, Delft, The Netherlands and its board size is 50 mm × 45 mm. The horizontal and vertical Interaction Off State On State field of view are 76° and 19°, and the minimum and maximum dis- tance for sensing are 1 m and 25 m. The sensor streams raw data to a PC via USB, with 2.5 W power consumption. We utilized the Button officially provided Matlab APIs to receive the streamed data and developed our detection algorithm, which will be described later 0.13 dB 0.83 dB in Section 7. Note in the equation below that the received power (𝑃𝑟𝑥 ), which is calculated through FFT computation on raw radar measurements, is linearly proportional to RCS of a target object (𝜎) Toggle at a fixed distance to the radar (𝑅), with a constant scale factor of Switch transmitted power level (𝑃𝑡𝑥 ), transmitter gain (𝐺𝑡𝑥 ), receiver gain -0.45 dB 3.23 dB (𝐺𝑟𝑥 ), and signal wavelength (𝜆): Figure 2: Two types of discrete controllers (top: button, bot- 𝑃𝑡𝑥 𝐺𝑡𝑥 𝐺𝑟𝑥 𝜎𝜆 2 tom: toggle switch). These controllers have two discrete 𝑃𝑟𝑥 = (4𝜋) 2 𝑅 4 states (i.e., on and off). As a result, we treated the received power as an indicator of RCS in the rest of this paper. • Knob (zigzag-fold): This controller works as a knob and has continuous states. It consists of a zigzag-fold and a 3d- 5 REFLECTOR DESIGN printed handle attached to the origami. The zigzag-fold part Our design rationale is to create shape-changing reflector struc- gradually opens as the handle is rotated in one direction and tures that can 1) be actuated by force at a single point, and 2) result closes as it is rotated in the opposite direction. in distinctive RCS at different shapes. To explore potential origami • Slider (Miura-fold): This controller works as a slider and has models suitable for our purpose, we first looked into a variety of continuous states. It consists of a Miura-fold [16] origami and existing works available on the Internet. We anticipated that struc- a 3d-printed handle. The Miura-fold part gradually expands tures consisting of mutually perpendicular surfaces would have and shrinks as the handle is manipulated by a user. high RCS values. This is inspired by the fact that corner reflectors, We measured RCS changes of each controller when they mor- which have three mutually perpendicular intersecting flat surfaces, phed. To do this, we first recorded signals without a controller as reflect waves directly towards the source, resulting in high RCS a base signal. Then, we placed a controller perpendicularly, 1 m values [12, 22]. away from the sensor. We recorded the signals with a controller We found four designs that could be used for our controllers. in each state and calculated the ratio of their amplitude to that of In general, all form-changing origami designs change their RCS the base signal, measured in dB. Overall, We found significant RCS when morphing. However, as we anticipated, the four selected changes between different states of the controllers. We elaborate an origami designs feature mutually perpendicular surfaces, which algorithm as to how these values are utilized for detection later in get distorted during the folding and unfolding process, resulting in Section 7 and document the evaluation of our system in Section 8. significant RCS changes. For example, we incorporated Miura-fold [16] into one controller that forms such perpendicular surfaces 6 USER INTERACTION when its structure is gradually expanded. Herein, we show the Before explaining the sensing algorithm, we describe our envisioned current controller designs: Button, Toggle Switch, Knob, and Slider future scenario of how users will set up and use our proposed (see Figure 2 and Figure 3). We used conventional silver origami, origami controllers (Figure 1). First, a user purchases controller tool paper with aluminum films coated on the surface, to further improve kits of interest from a distributor. These tool kits allow the user to the RF reflectance of surfaces. Our total material cost is less than easily make controllers from pieces on their own. Then, the user 40 cents per controller unit. opens a dedicated smartphone or smart speaker app to initiate the • Button (umbrella-fold): This controller works as a push-and- setup. The app will prompt the user to identify mm-wave sensors pull button and has discrete states (i.e., on and off). It consists in their room and origami controllers within the sensing range, of an umbrella-fold and a 3d-printed handle attached to the and start the calibration process. For calibration, he user collects origami. The umbrella part opens when the handle is pushed a small amount of sample data corresponding to each state of the and closes when it is pulled. controller, and the our system is ready to use. • Toggle Switch (corner reflector): This controller works as a toggle switch and has discrete states (i.e., on and off). It is a 7 SENSING ALGORITHM corner reflector structure made of cardboard papers coated Our detection approach is based on FMCW sensing [23]. At each with silver origami. The three surfaces get mutually per- frame, the Tx transmits 𝑁𝑐ℎ𝑖𝑟𝑝 chirps and there are 𝑁𝑠𝑎𝑚𝑝𝑙𝑒 sam- pendicular in the on state. This structure is disrupted when ples in each chirp, resulting in a reflected signal matrix: 𝑋𝑟𝑎𝑤 ∈ the switch is turned off. Users interact with the attached C𝑁𝑠𝑎𝑚𝑝𝑙𝑒 ×𝑁𝑐ℎ𝑖𝑟 𝑝 . When we apply fast Fourier transform (FFT) with 3d-printed handle to switch between the two states. the size of 𝑁 𝑓 𝑓 𝑡 , we get 𝑋 𝑓 𝑓 𝑡 ∈ C𝑁 𝑓 𝑓 𝑡 ×𝑁𝑐ℎ𝑖𝑟 𝑝 = 𝐹 𝐹𝑇 (𝑋𝑟𝑎𝑤 ). Then, CHIIoT 1, February 17, 2021, Delft, The Netherlands Riku Arakawa and Yang Zhang Interaction 0% 25 % 50 % 75 % 100 % Knob -0.22 dB 0.52 dB 0.89 dB 1.28 dB 1.42 dB Slider -0.49 dB 0.55 dB 0.83 dB 1.57 dB 2.29 dB Figure 3: Two types of continuous controllers (top: knob, bottom: slider). We expanded each controller from 0% to 100% with 25% interval. the largest value in each row (i.e., over 𝑁𝑐ℎ𝑖𝑟𝑝𝑠 chirps) was taken azimuth angle out after calculating the amplitude, resulting in a frame vector angle of incidence {0°, 15°, 30°, 45°} 𝑥 ∈ R𝑁 𝑓 𝑓 𝑡 ×1 = max |𝑋 𝑓 𝑓 𝑡 |. Each of the values in the 𝑁 𝑓 𝑓 𝑡 bins {0°, 15°, 30°, 45°} corresponds to the power of the reflected signal within specific 𝑐 , where mm-wave direction ranges from the sensor. The range resolution is given by 2𝐵 1m 𝑐 is the light speed and 𝐵 represents the used bandwidth. Our current implementation ignores frames that contain non- reflector norm negligible human body movements that interfere with our RCS room for testing test for angle of incidence sensing. To do this, we applied a simple threshold-based algorithm. In detail, we calculated the difference of two consecutive frames, say 𝑥𝑡 and 𝑥𝑡 +1 , and classified the frame 𝑥𝑡 +1 as containing human Figure 4: Setting of the pilot test for discrete detection. We body movements if the norm of the difference (i.e., |𝑥𝑡 +1 − 𝑥𝑡 |) is placed a controller to the gray positions and tested the detec- larger than a predefined threshold. Note that the current detection tion accuracy by changing its distance, azimuth angle, and algorithm requires users to exit the controller proximity after in- angle of incidence. teraction. This means our system only works in an asynchronous manner, where there could be a lag between user interaction and its detection. and linear, and thus our straightforward regression models worked Once a frame is detected as not containing human body move- sufficiently well in practice. ments, the frame is processed for detecting the controller’s state. Here, we assume that the system knows the bins of the feature 8 PILOT TEST vector 𝑥 that correspond to the controller position in terms of its We examined the accuracy of the proposed algorithm. In this pilot distance from the sensor. This information is provided during the test, we set the sensor parameters as followings: 𝑁𝑐ℎ𝑖𝑟𝑝 = 4 chirps user calibration process. By extracting the values in these bins of per frame, 𝑁𝑠𝑎𝑚𝑝𝑙𝑒 = 256 samples per chirp, 𝑁 𝑓 𝑓 𝑡 = 256, and 𝑥, we can focus on the data relevant to the controller’s state. Thus, 𝐵 = 200 MHz bandwidth. We averaged the measurements across we denote the values in these bins as a sub-vector 𝑦 and use it for the two Rxs for calculating 𝑥 and 𝑦 in this pilot test. We conducted the subsequent processing. the test in an open indoor space of approximately 10 𝑚 2 . For detecting discrete states of the controllers, users first collect data for on and off states of the controllers in the calibration, as we 8.1 Discrete Detection described in Section 6. Then, we calculated the mean values of data We first tested the accuracy of detecting discrete states of the con- collected from these two states as the thresholds to classify new trollers – button and toggle switch. frames of data after calibration. Similarly, for detecting continuous states of the controllers, users 8.1.1 Setting. We examined the sensing accuracy in a variety of set- provide sample data in the calibration for some of the data points in tings in terms of the controllers’ distance from the sensor, azimuth the detection range. We trained regression models (i.e., linear) with angle, and angle of incidence. Figure 4 (left) shows the locations the data collected during the user calibration process. Note that we we tested and Figure 4 (right) illustrates the angles of incidence found the RCS change as the controllers morph to be monotonous we tested. For each controller, we first placed it in the mm-wave Low-Cost Millimeter-Wave Interactive Sensing through Origami Reflectors CHIIoT 1, February 17, 2021, Delft, The Netherlands RF beam direction with 1 m interval up to 5 m. Then, we fixed the distance to 3 m and changed the azimuth angle with 15 degrees interval up to 45 degrees. Lastly, we fixed the distance to 3 m and the azimuth angle to 0 degrees, and changed the angle of incidence with 15 degrees interval up to 45 degrees. For each placement pattern, we first placed our controller and conducted calibration, as described in Section 7. After the cali- bration, we changed the state of the controller to be on and off repeatedly, five times each, while we recording the output of the algorithm each time. 8.1.2 Result. Figure 5 shows the accuracy, each corresponding to when we changed distance (top), azimuth angle (center), and angle of incidence (bottom). The toggle switch showed a stable detection accuracy over the conditions. On the other hand, the accuracy for detecting the button’s state gradually decreased as it was placed far from the sensor or its perpendicularity lost (i.e., as we changed the azimuth angle or angle of incidence). Overall, the high detection accuracy confirmed the validity of using our origami-based reflectors as discrete controllers. 8.2 Continuous Detection Next, we tested the performance of detecting continuous states of the controllers – knob and slider. 8.2.1 Setting. We first recorded signals without a controller placed in the environment and obtained 𝑦𝑏𝑎𝑠𝑒 . Then, we placed a controller 1 m away from the sensor board perpendicularly so that both the azimuth angle and angle of incidence were 0°. We then changed the expansion level of the controller from 0% to 100% with 25% interval, while we recording the corresponding signals 𝑦. We calculated and plotted the ratio of the amplitude of 𝑦 to 𝑦𝑏𝑎𝑠𝑒 in dB. We also trained a linear regression model and calculated the mean absolute percentage error (MAPE). 8.2.2 Result. Figure 6 shows the measured ratio of the amplitude of 𝑦 to 𝑦𝑏𝑎𝑠𝑒 in each of the controller’s expansion level. As expected, the values gradually increased as the controllers were expanded. The MAPE for each of the controllers are 33.2% (knob) and 28.6% (slider), respectively. The results clearly showed the correlation between the expansion level and RCS, with which our regression models can be easily trained for using origami-based reflectors as Figure 5: Accuracy of button and toggle switch when we continuous controllers. However, our proof-of-concept regressor changed their distance (top), azimuth angle (center), and an- implementation did not yield high accuracies due to large deviations gle of incidence (bottom) from the mm-wave sensor board. when controllers were expanded to certain levels (e.g., 75%). We suspect this issue was caused by fabrication defects, which we plan to further investigate and make improvements in our future work. DIY smart environment experience across a wide spectrum of ap- plications. 9 EXAMPLE APPLICATIONS As we discussed in Section 1, our low-cost controllers can be a 10 DISCUSSION AND FUTURE WORK promising approach for ubiquitous interactivity. For example, av- There are some directions to further refine our approach. First, erage homeowners can easily deploy the controllers into their en- the accuracy and robustness can be improved by adopting better vironments and connect them with various IoT applications, such fabrication process. For example, adding linings on top of basic as light, music player, TV, air conditioner, etc. Moreover, the con- origami structures could yield more programmable shape changing trollers can be installed in public places such as museums, hospitals, of the continuous controllers. We could also add protective coatings restaurants, and buses, replacing exiting controllers that are mostly to mitigate degradation of origami structures over time. wired and powered. Overall, we believe the advantages of our ap- Secondly, we will expand the origami design set. In this pa- proach being ultra-low-cost, wireless, and battery-free facilitate per, we demonstrated four designs, but considering the abundant CHIIoT 1, February 17, 2021, Delft, The Netherlands Riku Arakawa and Yang Zhang the same time, we would like to further investigate the interac- tion space where users can easily assemble controllers from basic material primitives, as we discussed in Section 2 and Section 6. 11 CONCLUSION In this paper, we proposed a novel approach to achieving ubiq- uitous interactivity: ultra-low-cost (less than 40 cents), wireless, battery-free controllers made of origami in concert with mm-wave sensing. We demonstrated four controller designs (i.e., button, tog- gle switch, knob, and slider). These controllers change their RCS significantly upon user interaction, which can be detected remotely by mm-wave sensing (e.g., FMCW). Our pilot test demonstrated the feasibility of the proposed approach. We believe that our work demonstrates a novel technique for ubiquitous interactivity, and will greatly facilitate users’ DIY of future smart environments. REFERENCES [1] Fadel Adib, Chen-Yu Hsu, Hongzi Mao, Dina Katabi, and Frédo Durand. 2015. Capturing the human figure through a wall. ACM Trans. Graph. 34, 6 (2015), 219:1–219:13. https://doi.org/10.1145/2816795.2818072 [2] Christopher Chen, David Howard, Steven L. Zhang, Youngwook Do, Sienna Sun, Tingyu Cheng, Zhong Lin Wang, Gregory D. Abowd, and HyunJoo Oh. 2020. SPIN (Self-powered Paper Interfaces): Bridging Triboelectric Nanogenerator with Folding Paper Creases. In TEI ’20: Fourteenth International Conference on Tangible, Embedded, and Embodied Interaction, Sydney, NSW, Australia, February 9-12, 2020. ACM, New York, NY, 431–442. https://doi.org/10.1145/3374920.3374946 [3] Jasper de Winkel, Vito Kortbeek, Josiah D. Hester, and Przemyslaw Pawelczak. 2020. Battery-Free Game Boy. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. 4, 3 (2020), 111:1–111:34. https://doi.org/10.1145/3411839 [4] EnOcean. [n.d.]. Product Information. Retrieved December 1, 2020 from Figure 6: The measured ratio of the amplitude of the re- https://www.enocean.com/en/products/battery-free-by-enocean/. [5] Mayank Goel, Chen Zhao, Ruth Vinisha, and Shwetak N. Patel. 2015. Tongue- flected signal to that of the base signal when controllers in-Cheek: Using Wireless Signals to Enable Non-Intrusive and Flexible Facial were expanded from 0% to 100% with 25% interval (top: knob, Gestures Detection. In Proceedings of the 33rd Annual ACM Conference on Human bottom: slider). Blue lines represent the fitted linear regres- Factors in Computing Systems, CHI 2015, Seoul, Republic of Korea, April 18-23, 2015. ACM, New York, NY, 255–258. https://doi.org/10.1145/2702123.2702591 sion model while orange dots correspond to the measured [6] Chen-Yu Hsu, Rumen Hristov, Guang-He Lee, Mingmin Zhao, and Dina Katabi. data points. 2019. Enabling Identification and Behavioral Sensing in Homes using Radio Reflections. In Proceedings of the 2019 CHI Conference on Human Factors in Com- puting Systems, CHI 2019, Glasgow, Scotland, UK, May 04-09, 2019. ACM, New York, NY, 548. https://doi.org/10.1145/3290605.3300778 [7] Ke Huo, Yuanzhi Cao, Sang Ho Yoon, Zhuangying Xu, Guiming Chen, and Karthik Ramani. 2018. Scenariot: Spatially Mapping Smart Things Within Augmented Reality Scenes. In Proceedings of the 2018 CHI Conference on Human Factors in origami structures found, we believe there are many other pos- Computing Systems, CHI 2018, Montreal, QC, Canada, April 21-26, 2018. ACM, New sible designs. We would like to run simulations to estimate RCS York, NY, 219. https://doi.org/10.1145/3173574.3173793 [8] Eugin Hyun and Jong-Hun Lee. 2009. A method for multi-target range and changes of origami structures for comprehensive exploration and velocity detection in automotive FMCW radar. In 2009 12th International IEEE optimization. Conference on Intelligent Transportation Systems. IEEE, New York, NY. https: Third, for achieving the ubiquitous interactivity through our //doi.org/10.1109/itsc.2009.5309873 [9] Vikram Iyer, Justin Chan, and Shyamnath Gollakota. 2017. 3D printing wireless controllers, it is demanded to enable the system to detect spatial connected objects. ACM Trans. Graph. 36, 6 (2017), 242:1–242:13. https://doi.org/ information of controllers. We would like to utilize beamforming 10.1145/3130800.3130822 [10] Runchang Kang, Anhong Guo, Gierad Laput, Yang Li, and Xiang ’Anthony’ Chen. and Angle of Arrival estimation with radars that have multiple 2019. Minuet: Multimodal Interaction with an Internet of Things. In Symposium transmitter and receiver antennas to increase spatial resolution. on Spatial User Interaction, SUI 2019, New Orleans, LA, USA, October 19-20, 2019. This improvement on spatial resolution will allow a user to deploy ACM, New York, NY, 2:1–2:10. https://doi.org/10.1145/3357251.3357581 [11] Mustafa Emre Karagozler, Ivan Poupyrev, Gary K. Fedder, and Yuri Suzuki. 2013. multiple controllers in the environment. Additionally, improved Paper generators: harvesting energy from touching, rubbing and sliding. In The spatial resolution will help locate users more precisely, which could 26th Annual ACM Symposium on User Interface Software and Technology, UIST’13, mitigate the current limitation that the users must be out of the St. Andrews, United Kingdom, October 8-11, 2013. ACM, New York, NY, 23–30. https://doi.org/10.1145/2501988.2502054 controller proximity for detection. [12] Eugene F Knott, John F Schaeffer, and Michael T Tulley. 2004. Radar cross section. Lastly, current controllers are relatively large in comparison SciTech Publishing. [13] Gierad Laput and Chris Harrison. 2019. SurfaceSight: A New Spin on Touch, User, to conventional ones. In future work, we will miniaturize the con- and Object Sensing for IoT Experiences. In Proceedings of the 2019 CHI Conference trollers with better fabrication techniques based on human-machine on Human Factors in Computing Systems, CHI 2019, Glasgow, Scotland, UK, May ergonomics. Specifically, we will apply automated fabrication tech- 04-09, 2019. ACM, New York, NY, 329. https://doi.org/10.1145/3290605.3300559 [14] Hanchuan Li, Eric Brockmeyer, Elizabeth J. Carter, Josh Fromm, Scott E. Hudson, niques such as laser cutting combined with vacuum forming, which Shwetak N. Patel, and Alanson P. Sample. 2016. PaperID: A Technique for will enable us to develop more complicated origami structures. At Drawing Functional Battery-Free Wireless Interfaces on Paper. In Proceedings of Low-Cost Millimeter-Wave Interactive Sensing through Origami Reflectors CHIIoT 1, February 17, 2021, Delft, The Netherlands the 2016 CHI Conference on Human Factors in Computing Systems, San Jose, CA, 1109/eumc.2007.4405477 USA, May 7-12, 2016. ACM, New York, NY, 5885–5896. https://doi.org/10.1145/ [27] Robert Xiao, Chris Harrison, and Scott E. Hudson. 2013. WorldKit: rapid and 2858036.2858249 easy creation of ad-hoc interactive applications on everyday surfaces. In 2013 [15] Jaime Lien, Nicholas Gillian, Mustafa Emre Karagozler, Patrick Amihood, Carsten ACM SIGCHI Conference on Human Factors in Computing Systems, CHI ’13, Paris, Schwesig, Erik Olson, Hakim Raja, and Ivan Poupyrev. 2016. Soli: ubiquitous France, April 27 - May 2, 2013. ACM, New York, NY, 879–888. https://doi.org/10. gesture sensing with millimeter wave radar. ACM Trans. Graph. 35, 4 (2016), 1145/2470654.2466113 142:1–142:19. https://doi.org/10.1145/2897824.2925953 [28] Hui-Shyong Yeo, Gergely Flamich, Patrick Schrempf, David Harris-Birtill, and [16] Koryo Miura and RJ Lang. 2009. The science of Miura-ori: A review. Origami 4 Aaron Quigley. 2016. RadarCat: Radar Categorization for Input & Interaction. (2009), 87–99. In Proceedings of the 29th Annual Symposium on User Interface Software and [17] Nintendo. [n.d.]. Product Information. Retrieved November 25, 2020 from Technology, UIST 2016, Tokyo, Japan, October 16-19, 2016. ACM, New York, NY, https://labo.nintendo.com. 833–841. https://doi.org/10.1145/2984511.2984515 [18] Md. Farhan Tasnim Oshim, Julian Killingback, Dave Follette, Huaishu Peng, and [29] Hui-Shyong Yeo, Ryosuke Minami, Kirill Rodriguez, George Shaker, and Aaron Tauhidur Rahman. 2020. MechanoBeat: Monitoring Interactions with Everyday Quigley. 2018. Exploring Tangible Interactions with Radar Sensing. Proc. ACM Objects using 3D Printed Harmonic Oscillators and Ultra-Wideband Radar. In Interact. Mob. Wearable Ubiquitous Technol. 2, 4 (2018), 200:1–200:25. https: UIST ’20: The 33rd Annual ACM Symposium on User Interface Software and Tech- //doi.org/10.1145/3287078 nology, Virtual Event, USA, October 20-23, 2020. ACM, New York, NY, 430–444. [30] Yang Zhang and Chris Harrison. 2018. Pulp Nonfiction: Low-Cost Touch Track- https://doi.org/10.1145/3379337.3415902 ing for Paper. In Proceedings of the 2018 CHI Conference on Human Factors in [19] Joseph A. Paradiso, Kai-yuh Hsiao, Joshua Strickon, Joshua Lifton, and Ari Adler. Computing Systems, CHI 2018, Montreal, QC, Canada, April 21-26, 2018. ACM, New 2000. Sensor systems for interactive surfaces. IBM Syst. J. 39, 3&4 (2000), 892–914. York, NY, 117. https://doi.org/10.1145/3173574.3173691 https://doi.org/10.1147/sj.393.0892 [31] Yang Zhang, Gierad Laput, and Chris Harrison. 2018. Vibrosight: Long-Range [20] Akarsh Prabhakara, Vaibhav Singh, Swarun Kumar, and Anthony Rowe. 2020. Vibrometry for Smart Environment Sensing. In The 31st Annual ACM Symposium Osprey: a mmWave approach to tire wear sensing. In MobiSys ’20: The 18th on User Interface Software and Technology, UIST 2018, Berlin, Germany, October 14- Annual International Conference on Mobile Systems, Applications, and Services, 17, 2018. ACM, New York, NY, 225–236. https://doi.org/10.1145/3242587.3242608 Toronto, Ontario, Canada, June 15-19, 2020. ACM, New York, NY, 28–41. https: [32] Yang Zhang, Chouchang (Jack) Yang, Scott E. Hudson, Chris Harrison, and //doi.org/10.1145/3386901.3389031 Alanson P. Sample. 2018. Wall++: Room-Scale Interactive and Context-Aware [21] QIACHIP. [n.d.]. Product Information. Retrieved December 1, 2020 from Sensing. In Proceedings of the 2018 CHI Conference on Human Factors in Computing https://qiachip.com/collections/qiachip-remote-control-switches. Systems, CHI 2018, Montreal, QC, Canada, April 21-26, 2018. ACM, New York, NY, [22] K. Sarabandi and Tsen-Chieh Chiu. 1996. Optimum corner reflectors for calibra- 273. https://doi.org/10.1145/3173574.3173847 tion of imaging radars. IEEE Transactions on Antennas and Propagation 44, 10 [33] Mingmin Zhao, Tianhong Li, Mohammad Abu Alsheikh, Yonglong Tian, Hang (1996), 1348–1361. https://doi.org/10.1109/8.537329 Zhao, Antonio Torralba, and Dina Katabi. 2018. Through-Wall Human Pose [23] Andrew G Stove. 1992. Linear FMCW radar techniques. In IEE Proceedings Estimation Using Radio Signals. In 2018 IEEE Conference on Computer Vision and F (Radar and Signal Processing), Vol. 139. IET, Institution of Engineering and Pattern Recognition, CVPR 2018, Salt Lake City, UT, USA, June 18-22, 2018. IEEE Technology (IET), 343–350. https://doi.org/10.1049/ip-f-2.1992.0048 Computer Society, New York, NY, 7356–7365. https://doi.org/10.1109/CVPR.2018. [24] Nicolas Villar and Steve Hodges. 2010. The peppermill: a human-powered user 00768 interface device. In Proceedings of the 4th International Conference on Tangible [34] Mingmin Zhao, Yingcheng Liu, Aniruddh Raghu, Hang Zhao, Tianhong Li, Anto- and Embedded Interaction 2010, Cambridge, MA, USA, January 24-27, 2010. ACM, nio Torralba, and Dina Katabi. 2019. Through-Wall Human Mesh Recovery Using New York, NY, 29–32. https://doi.org/10.1145/1709886.1709893 Radio Signals. In 2019 IEEE/CVF International Conference on Computer Vision, [25] Saiwen Wang, Jie Song, Jaime Lien, Ivan Poupyrev, and Otmar Hilliges. 2016. ICCV 2019, Seoul, Korea (South), October 27 - November 2, 2019. IEEE, New York, Interacting with Soli: Exploring Fine-Grained Dynamic Gesture Recognition in NY, 10112–10121. https://doi.org/10.1109/ICCV.2019.01021 the Radio-Frequency Spectrum. In Proceedings of the 29th Annual Symposium on [35] Yichao Zhao and Yi Su. 2017. Vehicles Detection in Complex Urban Scenes Using User Interface Software and Technology, UIST 2016, Tokyo, Japan, October 16-19, Gaussian Mixture Model With FMCW Radar. IEEE Sensors Journal 17, 18 (2017), 2016. ACM, New York, NY, 851–860. https://doi.org/10.1145/2984511.2984565 5948–5953. https://doi.org/10.1109/jsen.2017.2733223 [26] Volker Winkler. 2007. Range Doppler detection for automotive FMCW radars. In 2007 European Microwave Conference. IEEE, New York, NY. https://doi.org/10.