Search

Autonomous racing is becoming popular for academic and industry researchers as a test for general autonomous driving by pushing perception, planning, and control algorithms to their limits. While traditional control methods such as model predictive control are capable of generating an optimal control sequence at the edge of the vehicles’ physical controllability, these methods are sensitive to the accuracy of the modeling parameters, such as tire modeling coefficients. As model mismatch is inevitable in reality, the heuristic nature of Reinforcement Learning (RL) offers a viable approach to modeling robustness. This paper presents TC-Driver, an RL approach for robust control in autonomous racing. In particular, the TC-Driver agent is conditioned by a trajectory generated by any arbitrary traditional high-level trajectory planner. The proposed TC-Driver architecture addresses the tire parameter modeling inaccuracies by exploiting the learning capabilities of RL while utilizing the reliability of traditional planning methods in a hybrid fashion. We train the agent under varying tire conditions, allowing it to generalize to different model parameters, aiming to increase the racing capabilities of the system in practice. Experimental results demonstrate that the proposed hybrid RL architecture of the TC-Driver improves the generalization robustness of autonomous racing agents when compared to a previous state-of-the-art end-to-end-based architecture. Namely, the proposed controller yields a 29-fold improvement in crash ratio when facing model mismatch and can zero-shot transfer its behavior on unseen tracks which present completely new features, while the end-to-end baseline fails. When deployed on a physical system, the proposed architecture demonstrates zero-shot Sim2Real capabilities that outperform end-to-end agents 10-fold in terms of crash ratio while exhibiting similar driving characteristics in reality as in simulation., Field Robotics, 3, ISSN:2771-3989

In this work we identified the professional skills that students experience during PBL by analyzing the challenges they face and the resources they use during projects compared to exercise sessions. We showed that, overall, students experience more challenges during PBL compared to exercises. More specifically, more students experience difficulties designing solutions and making decisions during projects whereas more students experience difficulties expanding their skills during exercise sessions. We also showed that students most commonly deal with challenges by referring to class material during exercise sessions and gathering information during projects. Moreover, students experience a bigger variety of professional competencies during PBL compared to exercises. Finally, students are exposed to different challenges during the different project phases. During the problem definition phase, most students face challenges in engineering within constraints whereas, during the solution design phase, students have difficulties applying skills, designing solutions, and devising the process. During the problem implementation phase, most students encounter difficulties while applying knowledge and skills. Finally, the most challenging competency during the building, prototyping, and testing is interpreting data. While challenges vary, students use similar resources to deal with them across different stages.

Tactile sensing is a crucial perception mode for robots and human amputees in need of controlling a prosthetic device. Today robotic and prosthetic systems are still missing the important feature of accurate tactile sensing. This lack is mainly due to the fact that the existing tactile technologies have limited spatial and temporal resolution and are either expensive or not scalable. In this paper, we present the design and the implementation of a hardware-software embedded system called SmartHand. It is specifically designed to enable the acquisition and the real-time processing of high-resolution tactile information from a hand-shaped multi-sensor array for prosthetic and robotic applications. During data collection, our system can deliver a high throughput of 100 frames per second, which is 13.7x higher than previous related work. We collected a new tactile dataset while interacting with daily-life objects during five different sessions. We propose a compact yet accurate convolutional neural network that requires one order of magnitude less memory and 15.6x fewer computations compared to related work without degrading classification accuracy. The top-1 and top-3 cross-validation accuracies are respectively 98.86% and 99.83%. We further analyze the inter-session variability and obtain the best top-3 leave-one-out-validation accuracy of 77.84%. We deploy the trained model on a high-performance ARM Cortex-M7 microcontroller achieving an inference time of only 100 ms minimizing the response latency. The overall measured power consumption is 505 mW. Finally, we fabricate a new control sensor and perform additional experiments to provide analyses on sensor degradation and slip detection. This work is a step forward in giving robotic and prosthetic devices a sense of touch and demonstrates the practicality of a smart embedded system empowered by tiny machine learning., Comment: arXiv admin note: substantial text overlap with arXiv:2107.14598

Optical flow estimation is crucial for autonomous navigation and localization of unmanned aerial vehicles (UAV). On micro and nano UAVs, real-time calculation of the optical flow is run on low power and resource-constrained microcontroller units (MCUs). Thus, lightweight algorithms for optical flow have been proposed targeting real-time execution on traditional single-core MCUs. This paper introduces an efficient parallelization strategy for optical flow computation targeting new-generation multicore low power RISC-V based microcontroller units. Our approach enables higher frame rates at lower clock speeds. It has been implemented and evaluated on the eight-core cluster of a commercial octa-core MCU (GAP8) reaching a parallelization speedup factor of 7.21 allowing for a frame rate of 500 frames per second when running on a 50MHz clock frequency. The proposed parallel algorithm significantly boosts the camera frame rate on micro unmanned aerial vehicles, which enables higher flight speeds: the maximum flight speed can be doubled, while using less than a third of the clock frequency of previous singlecore implementations., 2022 IEEE International Symposium on Circuits and Systems (ISCAS), ISBN:978-1-6654-8485-5, ISBN:978-1-6654-8484-8, ISBN:978-1-6654-8486-2

Automatic gym activity recognition on energy- and resource-constrained wearable devices removes the human-interaction requirement during intense gym sessions - like soft-touch tapping and swiping. This work presents a tiny and highly accurate residual convolutional neural network that runs in milliwatt microcontrollers for automatic workouts classification. We evaluated the inference performance of the deep model with quantization on three resource-constrained devices: two microcontrollers with ARM-Cortex M4 and M7 core from ST Microelectronics, and a GAP8 system on chip, which is an open-sourced, multi-core RISC-V computing platform from GreenWaves Technologies. Experimental results show an accuracy of up to 90.4% for eleven workouts recognition with full precision inference. The paper also presents the trade-off performance of the resource-constrained system. While keeping the recognition accuracy (88.1%) with minimal loss, each inference takes only 3.2 ms on GAP8, benefiting from the 8 RISC-V cluster cores. We measured that it features an execution time that is 18.9x and 6.5x faster than the Cortex-M4 and Cortex-M7 cores, showing the feasibility of real-time on-board workouts recognition based on the described data set with 20 Hz sampling rate. The energy consumed for each inference on GAP8 is 0.41 mJ compared to 5.17 mJ on Cortex-M4 and 8.07 mJ on Cortex-M7 with the maximum clock. It can lead to longer battery life when the system is battery-operated. We also introduced an open data set composed of fifty sessions of eleven gym workouts collected from ten subjects that is publicly available.

Agriculture is the pillar industry for human survival. However, various diseases threaten the health of crops and lead to a decrease in yield. Industry 4.0 is making strides in plant illness prevention and detection, other than supporting farmers to improve plantations' income. To prevent crop diseases in time, this paper proposes, implements, and evaluates a low-power smart camera. It features a lightweight neural network to verify and monitor the growth status of crops. The proposed tiny model features optimized complexity, to be deployed in milliwatt power microcontrollers, and high accuracy. Experimental results show that our work reaches 99% accuracy on a 4-classes dataset and more than 96% for a 10 classes dataset. The compact model size (139 kB) and low complexity enable ultra-low power consumption (2.63 mW per hour) on the battery-powered Sony Spresense platform, which features a six-core ARM Cortex-M4F.

As the wind energy industry is maturing and wind turbines are becoming larger, there is an increasing need for cost-effective monitoring and data analysis solutions to understand the complex aerodynamic and acoustic behaviour of the flexible blades. Published measurements on operating rotor blades in real conditions are very scarce due to the complexity of the installation and use of measurement systems. However, recent developments in electronics, wireless communication and MEMS (micro-electromechanical systems) sensors are making it possible to acquire data in a cost-effective and energy-efficient way. In this work, therefore, a costeffective MEMS-based aerodynamic and acoustic wireless measurement system that is thin, non-intrusive, easy to install, low power and self-sustaining is designed and tested in a wind tunnel. The measurement system does not require an electrical connection to the wind turbine and can be mounted and removed without damaging the blade.The results show that the system is capable of delivering relevant results continuously, although work needs to be done on calibrating and correcting the pressure signals as well as on refining the concept for the attachment sleeve for weather protection in the field. Finally, two methods for using the measurements to provide added value to the wind energy industry are developed and demonstrated: (1) inferring the local angle of attack via stagnation point detection using differential pressure sensors near the leading edge and (2) detecting and classifying leading edge erosion using instantaneous snapshots of the measured pressure fields. Ongoing work involves field tests on a 6 kW operating wind turbine in Switzerland., Wind Energy Science, 7 (4)

Contrary to RGB cameras, Dynamic Vision Sensor (DVS) output visual data in the form of an asynchronous events stream by recording pixel-wise luminance changes at microsecond resolution. While conventional computer vision approaches utilise frame-based input data, thus failing to take full advantage of the high temporal resolution, novel approaches use spiking neural networks Spiking Neural Networks (SNNs) which are more compatible to handle event-based data since these bioinspired neural models intrinsically encode information in a sparse manner using activation spikes trains. This paper presents an attentional mechanism which detects regions with higher event density by using inherent SNN dynamics combined with online weight and threshold adaptation. We implemented the network directly on Intel's research neuromorphic chip Loihi and evaluate our proposed method on the open DVS128 Gesture Dataset. Our system is able to process 1 ms of event-data in 6 ms and reject more than 50% of incoming unwanted events occurring only 20 ms after activity onset.

The main motivation of paraglider pilots is to stay in the air for as long as possible. Therefore, paraglider pilots are always searching for thermal upwind that allow them to gain altitude. These thermal lifts are difficult to detect. Therefore, sensors and devices that indicate the vertical speed (so-called variometers) are widely used among paraglider pilots. This article presents the design and the implementation of an ultralow-power, self-sustaining, high-precision, wrist-worn variometer with a minimal form factor but infinite lifetime, which can visually and acoustically indicate the vertical velocity. This article demonstrates the benefits of combining multisource energy harvesting (EH) for wearable devices with low power design, exploiting a novel near-threshold ARM Cortex-M4F microcontroller, the Ambiq Apollo2, for the onboard processing. The experimental results show a power consumption of only $17.12~\mu \text{W}$ in sleep mode and $1937.21~\mu \text{W}$ in the worst case scenario when processing the data and outputting an audio feedback. Measurements confirmed that combining both thermal and solar EH makes the designed electronics self-sustaining. Without EH, the system will be operational for up to 372 h in always-on mode (worst case scenario) supplied by a 200-mAh lithium-ion battery.

Keyword spotting (KWS) is a crucial function enabling the interaction with the many ubiquitous smart devices in our surroundings, either activating them through wake-word or directly as a human-computer interface. For many applications, KWS is the entry point for our interactions with the device and, thus, an always-on workload. Many smart devices are mobile and their battery lifetime is heavily impacted by continuously running services. KWS and similar always-on services are thus the focus when optimizing the overall power consumption. This work addresses KWS energy-efficiency on low-cost microcontroller units (MCUs). We combine analog binary feature extraction with binary neural networks. By replacing the digital preprocessing with the proposed analog front-end, we show that the energy required for data acquisition and preprocessing can be reduced by 29x, cutting its share from a dominating 85% to a mere 16% of the overall energy consumption for our reference KWS application. Experimental evaluations on the Speech Commands Dataset show that the proposed system outperforms state-of-the-art accuracy and energy efficiency, respectively, by 1% and 4.3x on a 10-class dataset while providing a compelling accuracy-energy trade-off including a 2% accuracy drop for a 71x energy reduction.

Digital ultrasound (US) probes integrate the analog-to-digital conversion directly on the probe and can be conveniently connected to commodity devices. Existing digital probes are however limited to a relatively small number of channels, do not guarantee access to the raw US data, or cannot operate at very high frame rates (e.g., due to exhaustion of computing and storage units on the receiving device). In this work, we present an open, compact, power-efficient, 192-channels digital US data acquisition system capable of streaming US data at transfer rates greater than 80 Gbps towards a host PC for ultra-high frame rate imaging (in the multi-kHz range). Our US probe is equipped with two power-efficient Field Programmable Gate Arrays (FPGAs) and is interfaced to the host PC with two optical-link 100G Ethernet connections. The high-speed performance is enabled by implementing a Remote Direct Memory Access (RDMA) communication protocol between the probe and the controlling PC, that utilizes a high-performance Non-Volatile Memory Express (NVMe) interface to store the streamed data. To the best of our knowledge, thanks to the achieved datarates, this is the first high-channel-count compact digital US platform capable of raw data streaming at frame rates of 20 kHz (for imaging at 3.5 cm depths), without the need for sparse sampling, consuming less than 40 W.

The trend in Internet of Things research points toward performing increasingly compute-intensive data analysis tasks on embedded sensor nodes, rather than server centers. Exploiting the technological advances in both energy efficiency, and Tiny Machine Learning algorithms and methods, an increasing number of recognition and classification tasks can be performed by small, low-power, wireless sensor nodes. This paper presents WideVision, a wireless, wide-area sensing platform capable of performing on-board person detection with power requirements in the mW range. The WideVision platform integrates seamlessly into the Internet of Things, by coupling a dedicated multiradio platform, including a LoRa interface, enabling medium and long-range communication, with a novel parallel RISC-V microcontroller. We evaluate the proposed platform with the GAP8 microcontroller, which includes an 8-core RISC-V cluster, and greyscale camera to perform person detection by training and deploying an advanced, quantized neural network, achieving a statistical accuracy 84.5% for a 5-person detection task with a latency of only 182 ms. Experimental results demonstrate that the WideVision sensor node platform while performing inference at a rate of one image per minute on-board, is capable of lasting 300 days on a 2400 mAh Li-ion battery, and 65 days when evaluating one image per 10 seconds while providing effective surveillance of its perimeter., 2022 18th International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob), ISBN:978-1-6654-6975-3, ISBN:978-1-6654-6976-0

Long-term and maintenance-free operation is a critical feature for large-scale deployed battery-operated sensor nodes. Energy harvesting (EH) is the most promising technology to overcome the energy bottleneck of today's sensors and to enable the vision of perpetual operation. However, relying on fluctuating environmental energy requires an application-specific analysis of the energy statistics combined with an in-depth characterization of circuits and algorithms, making design and verification complex. This article presents a model-based design (MBD) approach for EH-enabled devices accounting for the dynamic behavior of components in the power generation, conversion, storage, and discharge paths. The extension of existing compact models combined with data-driven statistical modeling of harvesting circuits allows accurate offline analysis, verification, and validation. The presented approach facilitates application-specific optimization during the development phase and reliable long-term evaluation combined with environmental datasets. Experimental results demonstrate the accuracy and flexibility of this approach: the model verification of a solar-powered wireless sensor node shows a determination coefficient (R2) of 0.992, resulting in an energy error of only -1.57 % between measurement and simulation. Compared to state-of-practice methods, the MBD approach attains a reduction of the estimated state-of-charge error of up to 10.2 % in a real-world scenario. MBD offers non-trivial insights on critical design choices: the analysis of the storage element selection reveals a 2–3 times too high self-discharge per capacity ratio for supercapacitors and a peak current constrain for lithium-ion polymer batteries., Energy Conversion and Management, 272, ISSN:0196-8904, ISSN:1879-2227

Energy harvesting can enable wireless smart sensors to be self-sustainable by allowing them to gather energy from the environment. However, since the energy availability changes dynamically depending on the environment, it is difficult to find an optimal energy management strategy at design time. One existing approach to reflecting dynamic energy availability is energy-aware adaptive sampling, which changes the sampling rate of a sensor according to the energy state. This work proposes deep reinforcement learning-based predictive adaptive sampling for a wireless sensor node. The proposed approach applies deep reinforcement learning to find an effective adaptive sampling strategy based on the harvesting power and energy level. In addition, the proposed approach enables predictive adaptive sampling by designing adaptive sampling models that consider the trend of energy state. The evaluation results show that the predictive models can successfully manage the energy budget reflecting dynamic energy availability, maintaining a stable energy state for a up to 11.5% longer time., 2022 29th IEEE International Conference on Electronics, Circuits and Systems (ICECS), ISBN:978-1-6654-8823-5, ISBN:978-1-6654-8824-2

Indoor localization and context-awareness are becoming two of the key technologies for a large variety of applications. Real-time locating systems with centimeter accuracy and low power consumption have recently been made available by employing the Ultra WideBand (UWB) technology. Since 2015, Decawave has produced commercial UWB integrated circuits, exploiting time-of-flight measurement techniques to estimate the distance between two agents. This work presents a performance study between two Decawave transceivers, the DW1000 and the new DW3000 released in 2020. The testing space includes areas under line-of-sight and diverse non-line-of-sight conditions caused by the reflection of the UWB radio signals across various obstacles. Finally, we analyze the power consumption in distinct configurations, comparing the two devices. Results show that the two have similar precision in measurement ranges above one meter, while the DW3000 performs, on average, 33.2% better considering shorter distances. Moreover, the new transceiver features reduced power consumption by almost 50% during real-time measurements reaching an average value of 55 mW.

This article proposes a proof-of-concept device to continuously assess the usage of handheld power tools and detect construction working tasks (e.g., different drilling works) along with potential misusages, e.g., drops, with an energy-efficient architecture design. The designed device is based on Bluetooth low energy (BLE) and NFC connectivity. BLE is used to exchange data with a gateway, whereas NFC has been chosen as an energy-efficient wake-up mechanism. A temperature and humidity sensor is embedded to monitor storage conditions and an accelerometer for tool usage monitoring. The ARM Cortex-M4 core embedded in the BLE module is exploited to process the information at the edge. A Tiny Machine Learning (TinyML) algorithm is proposed to process the data directly on board and achieve low latency and high energy efficiency. The TinyML algorithm has been developed embedded in the proposed device to detect four different usage classes (tool transportation, no-load, metal, and wood drilling). A dataset containing more than 280 min of three-axis accelerations during different activities has been acquired with the device attached to a construction rotary hammer drill and used to train and validate the algorithm. A neural architecture search has been performed to optimize the trade-off between accuracy and complexity, achieving an accuracy of 90.6% with a model size of roughly 30 kB. The experimental results showed an ultralow power consumption in sleep mode of 550 nA and a peak power consumption of 8 mA while running TinyML on the edge. This results in a balanced combination of edge processing capabilities and low power consumption, enabling to obtain a smart Internet of Things (IoT) device in the field with a long lifetime of up to four years in operation and 17 years in shelf mode with a standard 250-mAh coin battery. This work enables a long battery lifetime operation of device degradation and utility analysis, further closing the gap between edge processing and fine granularity data evaluation., IEEE Transactions on Instrumentation and Measurement, 71, ISSN:0018-9456, ISSN:1557-9662

A challenging application scenario in the field of industrial Unmanned Aerial Vehicles (UAVs) is the capability of a robot to find and query smart sensor nodes deployed at arbitrary locations in the mission area. This work explores the combination of different communication technologies, namely, Ultra-Wideband (UWB) and Wake-Up Radio (WUR), with a UAV that acts as a “ubiquitous local-host” of a Wireless Sensor Network (WSN). First, the UAV performs the localization of the sensor node via multiple UWB range measurements, and then it flies in its proximity to perform energy-efficient data acquisition. We propose an energy-efficient and accurate localization algorithm – based on multi-lateration – that is computationally inexpensive and robust to in-field noise. Aiming at minimizing the sensor node energy consumption, we also present a communication protocol that leverages WUR technology to minimize ON-time of the power-hungry UWB transceiver on the sensors. In-field experimental evaluation demonstrates that our approach achieves a sub-meter localization precision of the sensor nodes – i.e., down to 0.6m – using only three range measurements, and runs in 4ms on a low power microcontroller (ARM Cortex-M4F). Due to the presence of the WUR and the proposed lightweight algorithm, the entire localization-acquisition cycle requires only 31mJ on the sensor node. The approach is suitable for several emerging Industrial Internet of Things application scenarios where a mobile vehicle needs to estimate the location of static objects without any precise knowledge of their position., Sustainable Computing: Informatics and Systems, 34

Radio Frequency (RF) wireless power transfer is a promising technology that has the potential to constantly power small Internet of Things (IoT) devices, enabling even battery-less systems and reducing their maintenance requirements. However, to achieve this ambitious goal, carefully designed RF energy harvesting (EH) systems are needed to minimize the conversion losses and the conversion efficiency of the limited power. For intelligent internet of things sensors and devices, which often have non-constant power requirements, an additional power management stage with energy storage is needed to temporarily provide a higher power output than the power being harvested. This paper proposes an RF wireless power energy conversion system for miniaturized IoT composed of an impedance matching network, a rectifier, and power management with energy storage. The proposed sub-system has been experimentally validated and achieved an overall power conversion efficiency (PCE) of over 30 % for an input power of −10 dBm and a peak efficiency of 57% at 3dBm., 2022 29th IEEE International Conference on Electronics, Circuits and Systems (ICECS), ISBN:978-1-6654-8823-5, ISBN:978-1-6654-8824-2

Ultrasound (US) is a widely used tool to observe musculoskeletal features during movement. US signals provide information about the instantaneous length and activation of muscles, and these signals are used for new prosthetic control strategies, to increase the performance of Human-Machine interfaces, or to improve muscle force predictions. However, existing commercial systems and research platforms are bulky, non-wearable, not supporting high frame rates, or lack configurability. Our work presents an US prototype meeting all key requirements to measure muscle activity during motion: our platform is open (providing access to raw data), flexible (operating in different modalities and channel count), it features wireless connectivity (supporting high frame rates> 100 Hz) and supports low power consumption (

The sophisticated sense of touch of the human hand significantly contributes to our ability to safely, efficiently, and dexterously manipulate arbitrary objects in our environment. Robotic and prosthetic devices lack refined tactile feedback from their end-effectors, leading to counterintuitive and complex control strategies. To address this lack, tactile sensors have been designed and developed, but they are either expensive and not scalable or offer an insufficient spatial and temporal resolution. This paper focuses on overcoming these issues by designing a smart embedded system, called SmartHand, enabling the acquisition and real-time processing of high-resolution tactile information from a hand-shaped multi-sensor array for prosthetic and robotic applications. We acquire a new tactile dataset consisting of 340,000 frames while interacting with 16 objects from everyday life and the empty hand, i.e., a total of 17 classes. The design of the embedded system minimizes response latency in classification, by deploying a small yet accurate convolutional neural network on a high-performance ARM Cortex-M7 microcontroller. Compared to related work, our model requires one order of magnitude less memory and 15.6× fewer computations while achieving similar inter-session accuracy and up to 98.86% and 99.83% top-1 and top-3 cross-validation accuracy, respectively. Experimental results of the designed prototype show a total power consumption of 505 mW and a latency of only 100 ms.

A nano-power integrated circuit is designed to convert the ac output of an electromagnetic micro generator for wearable kinetic energy harvesting to dc usable power. The converter is designed to minimize the quiescent current and to maximize the energy conversion and maximum power point tracking efficiency from the kinetic energy harvester. The architecture implements an ac/dc chopper exploiting the intrinsic inductance of the transducer combined with a zero-drop rectifier. An optimal switching scheme is derived for the converter and a corresponding control circuit is developed and optimized for low power consumption. The converter is designed in an SMIC 130-nm process, and circuit simulations show an efficiency of 87%-88% with respect to the maximum power point of the transducer with a battery supply voltage ranging from 3.5 V to 4.2 V. The quiescent current of the converter is as low as 250 nA.

Minimizing the power consumption of always-on sensors is crucial for extending the lifetime of battery-operated devices that are required to monitor events continuously and for long periods. This paper proposes a novel programmable $\mu \text{W}$ event-driven acoustic detector featuring “always-on” audio pattern recognition. The event-driven detector detects up to eight programmable spectral–temporal features extracted with a low-power single-channel analog circuit and classifies the features by an onboard microcontroller. The event-driven detector is combined with novel microbial fuel cells (MFCs) to achieve self-sustainability in an underwater scenario. Experimental results demonstrate that the power consumption of the detector is only $26.89~\mu \text{W}$ during always-on mode, achieving up to 59-dB sound pressure level of sensitivity. High detection accuracy of up to 95.89% in recognizing acoustic patterns has been experimentally verified. Accurate measurements with commercial MFCs demonstrate the capability to achieve self-sustainability in always-on monitoring.

Self-sustainability and intelligence will be critical features of the next generation of Internet of Things devices. Power management circuits need to be energy-efficient in multiple power modes, from nW sleep to mW active, and to support energy harvesting (EH) from multiple environmental sources. This article presents an adaptive and firmware-configurable power supply unit that enables intelligent devices to operate in the sub μW range while achieving self-sustainability through EH. The microcontroller-based design allows efficient conversion from batteries, dc, ac environmental sources, and nA quiescent currents. Source and load power points are decoupled with multiple dc–dc converters aiming to supply loads with adaptive voltage scaling and achieving high reliability. Experimental results show the flexibility and efficiency of this approach: the proposed power supply unit achieves a quiescent current of 54 nA and a maximum peak load current of 300 mA, delivered with an end-to-end power efficiency above 85%. © 2020 IEEE., IEEE Transactions on Power Electronics, 36 (5), ISSN:0885-8993, ISSN:1941-0107

The novel COVID-19 disease has been declared a pandemic event. Early detection of infection symptoms and contact tracing are playing a vital role in containing COVID-19 spread. As demonstrated by recent literature, multi-sensor and connected wearable devices might enable symptom detection and help tracing contacts, while also acquiring useful epidemiological information. This paper presents the design and implementation of a fully open-source wearable platform called H-Watch. It has been designed to include several sensors for COVID-19 early detection, multi-radio for wireless transmission and tracking, a microcontroller for processing data on-board, and finally, an energy harvester to extend the battery lifetime. Experimental results demonstrated only 5.9~mW of average power consumption, leading to a lifetime of 9 days on a small watch battery. Finally, all the hardware and the software, including a machine learning on MCU toolkit, are provided open-source, allowing the research community to build and use the H-Watch., 2021 IEEE International Symposium on Circuits and Systems (ISCAS), ISBN:978-1-7281-9201-7

Wearable, intelligent, and unobtrusive sensor nodes that monitor the human body and the surrounding environment have the potential to create valuable data for preventive human-centric ubiquitous healthcare. To attain this vision of unobtrusiveness, the smart devices have to gather and analyze data over long periods of time without the need for battery recharging or replacement. This article presents a software-configurable kinetic energy harvesting and power management circuit that enables self-sustainable wearable devices. By exploiting the kinetic transducer as an energy source and an activity sensor simultaneously, the proposed circuit provides highly efficient context-aware control features. Its mixed-signal nano-power context awareness allows reaching energy neutrality even in energy-drought periods, thus significantly relaxing the energy storage requirements. Furthermore, the asynchronous sensing approach also doubles as a coarse-grained human activity recognition frontend. Experimental results, using commercial micro-kinetic generators, demonstrate the flexibility and potential of this approach: the circuit achieves a quiescent current of 57nA and a maximum load current of 300mA, delivered with a harvesting efficiency of 79%. Based on empirically collected motion data, the system achieves an energy surplus of over 232mJ per day in a wrist-worn application while executing activity recognition at an accuracy of 89% and a latency of 60s., IEEE Transactions on Biomedical Circuits and Systems, 15 (5), ISSN:1932-4545, ISSN:1940-9990

Indoor Positioning System (IPS) is a crucial technology that enables medical staff and hospital managements to accurately locate and track persons or assets inside the medical buildings. Among other technologies, Bluetooth Low Energy (BLE) can be exploited for achieving an energy-efficient and low-cost solution. This work presents the design and implementation of an received signal strength indicator (RSSI)-based indoor localization system. The paper shows the implementation of a low complex weighted k-Nearest Neighbors algorithm that processes raw RSSI data from connection-less iBeacon’s. The designed hardware and firmware are implemented around the low-power and low-cost nRF52832 from Nordic Semiconductor. Experimental evaluation with the real-time data processing has been evaluated and presented in a 7.2 m by 7.2 m room with furniture and 5 beacon nodes. The experimental results show an average error of only 0.72 m in realistic conditions. Finally, the overall power consumption of the fixed beacon with a periodic advertisement of 100 ms is only 50 µA at 3 V, which leads to a long-lasting solution of over one year with a 500 mAh coin battery., 2021 17th International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob), ISBN:978-1-6654-2854-5, ISBN:978-1-6654-2855-2

Low-power energy harvesting has been demonstrated as a feasible alternative for the power supply of next-generation smart sensors and IoT end devices. In many cases, the output of kinetic energy harvesters is an alternating current (AC) requiring rectification in order to supply the electronic load. The rectifier design and selection can have a considerable influence on the energy harvesting system performance in terms of extracted output power and conversion losses. This paper presents a quantitative comparison of three passive rectifiers in a low-power, low-voltage electromagnetic energy harvesting sub-system, namely the full-wave bridge rectifier (FWR), the voltage doubler (VD), and the negative voltage converter rectifier (NVC). Based on a variable reluctance energy harvesting system, we investigate each of the rectifiers with respect to their performance and their effect on the overall energy extraction. We conduct experiments under the conditions of a low-speed rotational energy harvesting application with rotational speeds of 5 rpm to 20 rpm, and verify the experiments in an end-to-end energy harvesting evaluation. Two performance metrics—power conversion efficiency (PCE) and power extraction efficiency (PEE)—are obtained from the measurements to evaluate the performance of the system implementation adopting each of the rectifiers. The results show that the FWR with PEEs of 20% at 5 rpm to 40% at 20 rpm has a low performance in comparison to the VD (40–60%) and NVC (20–70%) rectifiers. The VD-based interface circuit demonstrates the best performance under low rotational speeds, whereas the NVC outperforms the VD at higher speeds (>18 rpm). Finally, the end-to-end system evaluation is conducted with a self-powered rpm sensing system, which demonstrates an improved performance with the VD rectifier implementation reaching the system’s maximum sampling rate (40 Hz) at a rotational speed of approximately 15.5 rpm., Sensors, 21 (18)

The mitigation of rapid mass movements involves a subtle interplay between field surveys, numerical modelling, and experience. Hazard engineers rely on a combination of best practices and, if available, historical facts as a vital prerequisite in establishing reproducible and accurate hazard zoning. Full-scale field tests have been performed to reinforce the physical understanding of debris flows and snow avalanches. Rockfall dynamics are - especially the quantification of energy dissipation during the complex rock-ground interaction - largely unknown. The awareness of rock shape dependence is growing, but presently, there exists little experimental basis on how rockfall hazard scales with rock mass, size, and shape. Here, we present a unique data set of induced single-block rockfall events comprising data from equant and wheel-shaped blocks with masses up to 2670 kg, quantifying the influence of rock shape and mass on lateral spreading and longitudinal runout and hence challenging common practices in rockfall hazard assessment., The awareness of rock shape dependence in rockfall hazard assessment is growing, but experimental and field studies are scarce. This study presents a large data set of induced single block rockfall events quantifying the influence of rock shape and mass on its complex kinematic behaviour.

This paper presents a low power, flexible and energy-efficient wireless sensor node for aerodynamic and acoustic measurements on wind turbine blades and other industrial structures. It comprises 40 high-accuracy absolute MEMS pressure sensors, ten MEMS microphones, a data processing system, a wireless transmitter based on Bluetooth Low Energy 5 tuned for long-range and high throughput while maintaining energy efficiency. The sensor node has been designed and implemented to test the range of communication, the impact on energy efficiency, the functionality, and the estimated lifetime. Experimental tests outdoor in realistic conditions revealed that the system can sustain a data rate of 850kbps over 438m. The node power consumption while streaming all measured data from a multi-MW wind turbine is only 46mW, enabling lifetimes of a full month even in the worst-case scenario of streaming all sensor data using an 8.7Ah Li-Ion battery.

LoRa is popular for internet of things applications as this communication technology offers both a long range and a low power consumption. However, LoRaWAN, the standard MAC protocol that uses LoRa as physical layer, has the bottleneck of a high downlink latency to achieve energy efficiency. To overcome this drawback we explore the use of wake-up radio combined with LoRa, and propose an adequate MAC protocol that takes profit of both these heterogeneous and complementary technologies. This protocol allows an opportunistic selection of a cluster head that forwards commands from the gateway to the nodes in the same cluster. Furthermore, to achieve self-sustainability, sensor nodes might include an energy harvesting sub-system, for instance to scavenge energy from the light, and their quality of service can be tuned, according to their available energy. To have an effective self-sustaining LoRa system, we propose a new energy manager that allows less fluctuations of the quality of service between days and nights. Latency and energy are modeled in a hybrid manner, i.e., leveraging microbenchmarks on real hardware platforms, to explore the influence of the energy harvesting conditions on the quality of service of this heterogeneous network. It is clearly demonstrated that the cooperation of nodes within a cluster drastically reduces the latency of LoRa base station commands, e.g., by almost 90% compared to traditional LoRa scheme for a 10 nodes cluster. © 2021 by the authors. Licensee MDPI, Basel, Switzerland., Sensors, 21 (3), ISSN:1424-8220

Predictive maintenance and structural health monitoring are challenging and promising research fields today. In particular, cost-effective and long-term monitoring of wind turbines has been proven to be one of the key elements to successfully increase their efficiency. Accurate numerical modeling and real-time control-in-the-loop play an increasingly prominent role in understanding and optimizing blade aerodynamic and acoustic performances. A non-intrusive and modular measurement system is a prerequisite for long-term measurement campaigns in existing and future wind turbines. Current methods of performing aerodynamic and acoustic field measurements are cumbersome and expensive, leading to a shortage of aerodynamic and acoustic datasets on operating wind turbines. This paper demonstrates the ability of the new Aerosense system to operate successfully in the field. Aerosense is a long-lasting battery-operated and flexible wireless sensor node that can directly measure aerodynamic and acoustic effects on wind turbine blades. It consists of an array of state-of-the-art Micro-Electro-Mechanical Systems (MEMS) sensors, including 40 barometers and 10 microphones, combined with an ultra low power system-on-chip with wireless transmission over Bluetooth 5.1. Experimental results demonstrate the possibility of continuously acquiring data for up to four months on a single lithium battery of 8.7 Ah, featuring an absolute accuracy of 10Pa and an audio bandwidth of 6kHz., Journal of Physics: Conference Series, 2265 (2), ISSN:1742-6588, ISSN:1742-6596

Sound event detection (SED) is a hot topic in consumer and smart city applications. Existing approaches based on Deep Neural Networks are very effective, but highly demanding in terms of memory, power, and throughput when targeting ultra-low power always-on devices. Latency, availability, cost, and privacy requirements are pushing recent IoT systems to process the data on the node, close to the sensor, with a very limited energy supply, and tight constraints on the memory size and processing capabilities precluding to run state-of-the-art DNNs. In this paper, we explore the combination of extreme quantization to a small-footprint binary neural network (BNN) with the highly energy-efficient, RISC-V-based (8+1)-core GAP8 microcontroller. Starting from an existing CNN for SED whose footprint (815 kB) exceeds the 512 kB of memory available on our platform, we retrain the network using binary filters and activations to match these memory constraints. (Fully) binary neural networks come with a natural drop in accuracy of 12-18% on the challenging ImageNet object recognition challenge compared to their equivalent full-precision baselines. This BNN reaches a 77.9% accuracy, just 7% lower than the full-precision version, with 58 kB (7.2 times less) for the weights and 262 kB (2.4 times less) memory in total. With our BNN implementation, we reach a peak throughput of 4.6 GMAC/s and 1.5 GMAC/s over the full network, including preprocessing with Mel bins, which corresponds to an efficiency of 67.1 GMAC/s/W and 31.3 GMAC/s/W, respectively. Compared to the performance of an ARM Cortex-M4 implementation, our system has a 10.3 times faster execution time and a 51.1 times higher energy-efficiency., Comment: 6 pages conference

2020 IEEE 18th International Conference on Industrial Informatics (INDIN), ISBN:978-1-7281-4964-6

This work proposes a low-power high-accuracy embedded hand-gesture recognition algorithm targeting battery-operated wearable devices using low-power short-range RADAR sensors. A 2-D convolutional neural network (CNN) using range-frequency Doppler features is combined with a temporal convolutional neural network (TCN) for time sequence prediction. The final algorithm has a model size of only 46 thousand parameters, yielding a memory footprint of only 92 KB. Two data sets containing 11 challenging hand gestures performed by 26 different people have been recorded containing a total of 20’210 gesture instances. On the 11 hand gesture data set, accuracies of 86.6% (26 users) and 92.4% (single user) have been achieved, which are comparable to the state of the art, which achieves 87% (10 users) and 94% (single user), while using a TCN-based network that is $7500\times $ smaller than the state of the art. Furthermore, the gesture recognition classifier has been implemented on a parallel ultralow power processor, demonstrating that real-time prediction is feasible with only 21 mW of power consumption for the full TCN sequence prediction network, while a system-level power consumption of less than 120 mW is achieved. We provide open-source access to example code and all data collected and used in this work on tinyradar.ethz.ch.

This paper presents an accurate and robust embedded motor-imagery brain–computer interface (MI-BCI). The proposed novel model, based on EEGNet [1], matches the requirements of memory footprint and computational resources of low-power microcontroller units (MCUs), such as the ARM Cortex-M family. Furthermore, the paper presents a set of methods, including temporal downsampling, channel selection, and narrowing of the classification window, to further scale down the model to relax memory requirements with negligible accuracy degradation. Experimental results on the Physionet EEG Motor Movement/Imagery Dataset show that standard EEGNet achieves 82.43%, 75.07%, and 65.07% classification accuracy on 2-, 3-, and 4-class MI tasks in global validation, outperforming the state-of-the-art (SoA) convolutional neural network (CNN) by 2.05%, 5.25%, and 6.49%. Our novel method further scales down the standard EEGNet at a negligible accuracy loss of 0.31% with 7.6× memory footprint reduction and a small accuracy loss of 2.51% with 15× reduction. The scaled models are deployed on a commercial Cortex-M4F MCU taking 101 ms and consuming 4.28 mJ per inference for operating the smallest model, and on a Cortex-M7 with 44 ms and 18.1 mJ per inference for the medium-sized model, enabling a fully autonomous, wearable, and accurate low-power BCI. © 2020 IEEE., 2020 IEEE International Symposium on Medical Measurements and Applications (MeMeA), ISBN:978-1-7281-5386-5, ISBN:978-1-7281-5387-2