Plastics recycling is a critical ecological and economic solution to manage plastic waste, yet a staggering proportion of plastics from daily use is landfilled or incinerated. A critical step to recycling plastics is our ability to sort plastics by type (HDPE, LDPE, PET, PP, PS, and PVC) at a mixed recycling facility. However, challenges such as sensor system cost, difficulty in data collection, and the dense sampling required for model fine-tuning continue to hinder reliable large-scale deployment and limit progress toward a sustainable circular plastic economy. In this work, we propose a novel physics-informed plastics classification system based on active thermal imaging. Additionally, we present a two-stage training strategy that uses a large quantity of easily generated PDE-based simulated samples for pretraining and fine-tunes the model to real-world data distributions using only a sparse set of samples. We validate the efficacy of the proposed approach on real-world plastic samples. Thus, we introduce Thermal Imaging based Plastic Sorting (TIPS), a system that achieves up to 100% and 94.7% accuracy in plastic type classification for black and white plastics, respectively.

The smart buildings of the future need to understand the movement and occupancy of the people in the environment. Using cameras to provide this context can be privacy-invasive. Alternatively, installing dedicated hardware to sense the environment can be cost-prohibitive and limit ubiquitous adoption. WiFi-based sensing has hence been championed to provide this building-scale sensing, as it allows for both privacy and is ubiquitously deployed in most buildings. However, industry-translatable research in this space has been challenging as no building-scale systems can provide WiFi sensing data. Consequently, many real-world challenges of deploying these sensing systems remain a mystery. To overcome this veil of mystery, we develop and open-source WiSenseHub, a building-scale WiFi-sensing system. We build our system on commercially available WiFi radios, deploy our backend services to collect data on infinitely scalable AWS cloud or a local server desktop, and build a front-end phone-based interface to collect diverse WiFi sensing data. We deployed multiple WiFi radios in our building and collected data for user devices for over 38 hours.

Interest in autonomous navigation and exploration for indoor applications has spurred research into indoor Simultaneous Localization and Mapping (SLAM) robot systems. While most of these SLAM systems use camera and LiDAR sensors in tandem with an odometry sensor, these odometry sensors drift over time. Visual (LiDAR/camera-based) SLAM systems deploy compute and memory-intensive search algorithms to detect 'Loop Closures' to combat this drift, making the trajectory estimate globally consistent. Instead, WAIS (WiFi Assisted Indoor SLAM) demonstrates using WiFi-based sensing can reduce this resource intensiveness drastically. By covering over 1500 m in realistic indoor environments and WiFi deployments, we showcase 4.3× and 4× reduction in compute and memory consumption compared to state-of-the-art Visual and Lidar SLAM systems. Incorporating WiFi into the sensor stack improves the resiliency of the Visual-SLAM system. We find the 90th percentile translation errors improve by ~ 40% and orientation errors by ~ 60% compared with purely camera-based systems. Additionally, we open-source a toolbox, WiROS, to furnish online and compute efficient WiFi measurements.

Wide-scale adoption of VR/AR technologies in gaming, video conferencing, and for other remote telepresence applications demands limb tracking for a more immersive experience. In an attempt to bolster limb tracking, we present UWBTrac, a UWB + IMU based fusion tracker for VR applications. In this demo, and accompanying video1, we showcase this UWB tracker in comparison with HTC Vive VR trackers.

This paper presents SSLIDE, Sound Source Localization for Indoors using DEep learning, which applies deep neural networks (DNNs) with encoder-decoder structure to localize sound sources with random positions in a continuous space. The spatial features of sound signals received by each microphone are extracted and represented as likelihood surfaces for the sound source locations in each point. Our DNN consists of an encoder network followed by two decoders. The encoder obtains a compressed representation of the input likelihoods. One decoder resolves the multipath caused by reverberation, and the other decoder estimates the source location. Experiments based on both the simulated and experimental data show that our method can not only outperform multiple signal classification (MUSIC), steered response power with phase transform (SRP-PHAT), sparse Bayesian learning (SBL), and a competing convolutional neural network (CNN) approach in the reverberant environment but also achieve a good generalization performance.

Location services, fundamentally, rely on two components: a mapping system and a positioning system. The mapping system provides the physical map of the space, and the positioning system identifies the position within the map. Outdoor location services have thrived over the last couple of decades because of well-established platforms for both these components (e.g. Google Maps for mapping, and GPS for positioning). In contrast, indoor location services haven't caught up because of the lack of reliable mapping and positioning frameworks. Wi-Fi positioning lacks maps and is also prone to environmental errors. In this paper, we present DLoc, a Deep Learning based wireless localization algorithm that can overcome traditional limitations of RF-based localization approaches (like multipath, occlusions, etc.). We augment DLoc with an automated mapping platform, MapFind. MapFind constructs location-tagged maps of the environment and generates training data for DLoc. Together, they allow off-the-shelf Wi-Fi devices like smartphones to access a map of the environment and to estimate their position with respect to that map. During our evaluation, MapFind has collected location estimates of over 105 thousand points under 8 different scenarios with varying furniture positions and people motion across two different spaces covering 2000 sq. Ft. DLoc outperforms state-of-the-art methods in Wi-Fi-based localization by 80% (median & 90th percentile) across the two different spaces.

Bluetooth Low Energy (BLE) tags have become very prevalent over the last decade for tracking applications in homes as well as businesses. These tags are used to track objects, navigate people, and deliver contextual advertisements. However, in spite of the wide interest in tracking BLE tags, the primary methods of tracking them are based on signal strength (RSSI) measurements. Past work has shown that such methods are inaccurate, and prone to multipath and dynamic environments. As a result, localization using Wi-Fi has moved to Channel State Information (CSI, includes both signal strength and signal phase) based localization methods. In this paper, we seek to investigate what are the challenges that prevent BLE from adopting CSI based localization methods. We identify fundamental differences at the PHY layer between BLE and Wi-Fi, that make it challenging to extend CSI based localization to BLE. We present our system, BLoc, that incorporates novel, BLE-compatible algorithms to overcome these challenges and enable an accurate, multipath-resistant localization system. Our empirical evaluation shows that BLoc can achieve a localization accuracy of 86 cm with BLE tags, a 3X improvement over a state-of-the-art baseline.