The presence of concept drift poses challenges for anomaly detection in time series. While anomalies are caused by undesirable changes in the data, differentiating abnormal changes from varying normal behaviours is difficult due to differing frequencies of occurrence, varying time intervals when normal patterns occur, and identifying similarity thresholds to separate the boundary between normal vs. abnormal sequences. Differentiating between concept drift and anomalies is critical for accurate analysis as studies have shown that the compounding effects of error propagation in downstream data analysis tasks lead to lower detection accuracy and increased overhead due to unnecessary model updates. Unfortunately, existing work has largely explored anomaly detection and concept drift detection in isolation. We develop AnDri, a system for Anomaly detection in the presence of Drift, and enables users to interactively co-explore the interaction of anomalies and drift. Our system demonstration provides two motivating scenarios that extend existing anomaly detection baselines with partial labels towards improved co-detection accuracy, and highlights the superiority of AnDri over these baselines.

This work presents a high-speed 3D Gaussian Splatting-based SLAM (3DGS-SLAM) accelerator to support dense mapping for mobile devices. 3DGS-SLAM has two main hardware challenges for acceleration: 1) Large α-computation computation. 2) Memory bottleneck caused by irregular memory access and large number of Gaussians. First, diagonal feeding (DF) controller precludes redundant-α computation, and symmetric alpha reuse (SAR) enables reusing computed alpha. This method reduces 35.3% system computation. Second, voxel-based inter-frame caching (VIFC) enables selective inter-frame voxel caching, which reduces 44.0% of external memory access. As a result, the proposed 3DGS-SLAM accelerator achieves 51.2 fps with 0.07µJ/point with support voltage 0.9V, clock frequency 200MHz mapping a high-quality dense-map.

An energy-efficient semantic segmentation (SS) processor, achieving 3.55 mJ/frame system energy efficiency, is proposed. To address the challenges posed by Mixed Transformer (MiT)-based SS, including high external memory bandwidth requirement and large on-chip memory footprint, we introduce a novel compression method called Chunk-based Bit Plane Compression (CBPC). CBPC leverages the high inter-token locality of feature maps in MiT-based SS, along with the robustness and compression ratio variations based on bit position to achieve a high compression ratio. To support CBPC, we propose an area and power-efficient CBPC encoder/decoder. In addition, a Similar Token Coarse Skipping (STCS) Core is proposed for high throughput. It enables row-wise clock gating and array-wise coarse skipping to reduce redundant computation. By removing redundant computation, the processor achieves higher throughput and lower computation power. The proposed processor reduces 67.6% of EMA power and accomplishes 19.24 TOPS/W core energy efficiency. The proposed processor achieves 44.3% higher system energy efficiency than the previous processors.

A novel energy-efficient semantic segmentation (SS) processor is proposed for achieving high system energy efficiency on mobile devices. 1) Excessive external memory access and 2) a large amount of redundant computation hinders energy-efficient SS acceleration. Three key features enable real-time energy-efficient CNN/ViT hybrid SS. A new compression method named Chunk-based Bit Plane Compression (CBPC) reduces the memory footprint and energy consumption due to external memory access. CBPC enhances compression ratio by leveraging the high inter-token similarity of feature maps and applying bit plane compression in sign-magnitude data representation, using chunk-wise low-bit plane shared bias. The proposed CBPC encoder/decoder supports CBPC with minimum area overhead. Additionally, the Similar Token Coarse Skipping (STCS) Core enhances the throughput and reduces the computation power by eliminating redundant computations. STCS core employs Row-wise Line Gating for low-power computation and Array-wise Coarse Skipping to minimize redundant computation. As a result, our proposed processor reduces external memory access energy by 67.6% and achieves a core energy efficiency of 19.24 TOPS/W. Our solution achieves 3.55mJ/frame system-level energy efficiency which is 79.7% higher than the previous SOTA SS processor.

Efficient commercial airline operations rely extensively on consistent and timely flight departure and arrival times, adherence to regular maintenance schedules, and minimizing unexpected service interruptions. Such interruptions include mechanical malfunctions that require unscheduled removal and replacement of a part or component. This necessarily causes cascading delays, consequential customer dissatisfaction, and increased costs. In this paper, we propose a novel clustering-based framework that takes historical (scheduled and unscheduled) maintenance events for aircraft, to predict when the next maintenance event will occur. We identify clusters of aircraft sharing similar spatio-temporal performance patterns, and define a prediction model over each cluster. We show that our models achieve improved accuracy over naive and baseline approaches, under varying parameters, airline carriers, and aircraft part components.

RPL is an IPv6 routing protocol for low-power and lossy networks (LLNs) designed to meet the requirements of a wide range of LLN applications including smart grid AMIs, home and building automation, industrial and environmental monitoring, health care, wireless sensor networks, and the Internet of Things (IoT) in general with thousands and millions of nodes interconnected through multihop mesh networks. RPL constructs tree-like routing topology rooted at an LLN border router (LBR) and supports bidirectional IPv6 communication to and from the mesh devices by providing both upward and downward routing over the routing tree. In this article, we focus on the interoperability of downward routing and supporting its two modes of operations (MOPs) defined in the RPL standard (RFC 6550). Specifically, we show that there exists a serious connectivity problem in RPL protocol when two MOPs are mixed within a single network, even for standard-compliant implementations, which may result in network partitions. To address this problem, this article proposes DualMOP-RPL , an enhanced version of RPL, which supports nodes with different MOPs for downward routing to communicate gracefully in a single RPL network while preserving the high bidirectional data delivery performance. DualMOP-RPL allows multiple overlapping RPL networks in the same geographical regions to cooperate as a single densely connected network even if those networks are using different MOPs. This will not only improve the link qualities and routing performances of the networks but also allow for network migrations and alternate routing in the case of LBR failures. We evaluate DualMOP-RPL through extensive simulations and testbed experiments and show that our proposal eliminates all the problems we have identified.

The capability to extract human activity context in a room environment can be used as meaningful feedback for various wireless indoor application systems. Being able to do so with easily installable resource-limited sensing components can even further increase the system's applicability for various purposes. This paper introduces our efforts to design a system consisting of heterogeneous low-cost, resource-limited, wireless sensing platforms for accurately extracting the human activity context from an indoor environment. Specifically, we introduce Resource Limited Sensor-based activity Context Extraction (ReLiSCE), a system consisting of microphone array, passive infra-red (PIR), and illumination sensors that effectively detect the activities that occur in an office (meeting room) environment. The signal processing schemes used in ReLiSCE are designed so that their size and complexity is suitable for the resource limitations that many embedded computing platforms introduce. Using empirical evaluations with a prototype system, we show that despite the simplicity of its data processing schemes, ReLiSCE successfully classifies human activity states in various meeting scenarios. Furthermore, we show that high accuracy is achieved by combining results from heterogeneous sensors. We foresee this paper as a sub-system that interconnects with various application systems for autonomously configuring people's everyday living environments in a more comfortable and energy-efficient manner.

Tightly synchronizing transmissions of the same packet from different sources theoretically results in constructive interference. Exploiting this property potentially speeds up network-wide packet propagation with minimal latencies. Our empirical results suggest the timing constraints can be relaxed in the real world, especially for radios using lower frequencies such as the IEEE 802.15.4 radios at 900 MHz. Based on these observations we propose PEASST, a topology-free protocol that leverages synchronized transmissions to lower the cost of end-to-end data transfers, and enables multiple traffic flows. In addition, PEASST integrates a receiver-initiated duty-cycling mechanism to further reduce node energy consumption. Results from both our Matlab-based simulations and indoor testbed reveal that PEASST can achieve a packet delivery latency matching the current state-of-the-art schemes that also leverages synchronized transmissions. In addition, PEASST reduces the radio duty-cycling by three-fold. Furthermore, comparisons with a multi-hop routing protocol shows that PEASST effectively reduces the per-packet control overhead. This translates to a ~10% higher packet delivery performance with a duty cycle of less than half.

In this work we point out the issue of the IETF RPL routing protocol's two different downwards routing schemes not being able to interoperate with each other. This problem is less of an issue when low-power and lossy networks (LLNs) are deployed homogeneously but with the industrial kickoff and large scale deployments, the interoperability of heterogeneous, standards-compliant implementations will become a significant issue. To address this, we suggest two major changes to IETF RPL (RFC 6550). First we suggest that all storing mode nodes should hold the capability to understand and attach source routing headers that the non-storing mode nodes require to forward packets. Next, we suggest that RPL's non-storing mode nodes should send their destination advertisement messages hop-by-hop, rather than the current end-to-end approach. We show, with two different IPv6 implementations in TinyOS and NanoQplus, that our suggestions high achieve high interoperability performance among different implementations for downwards traffic patterns.

As the number of sensors increase in wireless sensing applications, it is important for nodes to provide meaningful summary reports of the original data to the gateway. In doing so, given the resource constraints of the sensing devices, we need a light weight, yet, effective scheme to minimize the number of reports at the sensors while preserving the accuracy of the original data. However, we show in this work that unlike outdoors environments where various sensors may show a similar phenomena (e.g., high spatial correlation), this may not be true for sensors deployed in a typical indoors environment. To resolve this issue, we introduce a data summarizing scheme for such indoor applications that combines two techniques. First, our scheme detects events in a data stream by comparing the short term mean of the recent measurements with the most recent report sent to the gateway. Second, we include an exponentially increasing/decreasing timer that triggers additional reports where the timer's interval is reconfigured dynamically with respect to the result of our event detection method. Evaluations with temperature and humidity data collected in an indoors environment indicate that our scheme significantly reduces the number of transmissions while maintaining a mean error as low as ~0.07°C and ~0.08%RH.