REAP: Runtime Energy-Accuracy Optimization for Energy Harvesting IoT Devices

  title={REAP: Runtime Energy-Accuracy Optimization for Energy Harvesting IoT Devices},
  author={Ganapati Bhat and Kunal Bagewadi and Hyung Gyu Lee and {\"U}mit Y. Ogras},
  journal={2019 56th ACM/IEEE Design Automation Conference (DAC)},
The use of wearable and mobile devices for health and activity monitoring is growing rapidly. These devices need to maximize their accuracy and active time under a tight energy budget imposed by battery and form-factor constraints. This paper considers energy harvesting devices that run on a limited energy budget to recognize user activities over a given period. We propose a technique to co-optimize the accuracy and active time by utilizing multiple design points with different energy-accuracy… 

Figures and Tables from this paper

End-to-End Approximation for Characterizing Energy Efficiency of IoT Applications

The authors' analysis showed variable significance in terms of energy gains upon approximating applications at different phases, which span over a range of up to 25% gains at sense phase, 88% in the compute phase and 67% at the transmit phase.

Adaptive Energy Management for Self-Sustainable Wearables in Mobile Health

This paper studies the novel problem of adaptive energy management towards the goal of self-sustainable wearables by using harvested energy to supplement the battery energy and to reduce manual recharging by users using a principled algorithm referred as AdaEM.

An Ultra-Low Energy Human Activity Recognition Accelerator for Wearable Health Applications

This work presents the first fully integrated custom hardware accelerator (HAR engine) that consumes 22.4 μJ per operation using a commercial 65 nm technology and achieves 95% accuracy in recognizing 8 common human activities while providing three orders of magnitude higher energy efficiency compared to existing solutions.

Highly Adaptive Linear Actor-Critic for Lightweight Energy-Harvesting IoT Applications

The LAC-AB algorithm is proposed that introduces into the LAC algorithm an adaptive learning rate called Adam for actor update to achieve better adaptability and a definition of “convergence” when quantitative analysis of convergence is performed is introduced.

ECO: Enabling Energy-Neutral IoT Devices Through Runtime Allocation of Harvested Energy

This work presents a runtime energy-allocation framework that optimizes the utility of the target device under energy constraints using a rollout algorithm, which is a sequential approach to solve dynamic optimization problems.

AdaSense: Adaptive Low-Power Sensing and Activity Recognition for Wearable Devices

This paper presents AdaSense: a sensing, feature extraction and classification co-optimized framework for Human Activity Recognition that achieves 69% reduction in the power consumption of the sensor with less than 1.5% decrease in the activity recognition accuracy.



Near-optimal energy allocation for self-powered wearable systems

This paper presents a near-optimal runtime energy management technique by considering the harvested energy and shows that the results of the proposed algorithm are, on average, within 3% of the optimal solution computed offline.

Adaptive Control of Duty Cycling in Energy-Harvesting Wireless Sensor Networks

A new technique for solving the problem of adaptive duty-cycling based on results from adaptive control theory is introduced and it is shown that it achieves better performance than previous approaches on a broader class of energy source data sets.

Energy harvesting in wireless sensor networks: A comprehensive review

Power management in energy harvesting sensor networks

Power management is an important concern in sensor networks, because a tethered energy infrastructure is usually not available and an obvious concern is to use the available battery energy

Powering the Internet of Things

The design of ultra-low power hardware platforms that integrate computing, sensing, storage, and wireless connectivity in a tiny form factor, the development of intelligent system-level power management techniques, and the use of environmental energy harvesting to make IoT devices self-powered, thus decreasing - in some cases, even eliminating - their dependence on batteries are highlighted.

Dynamic power management for long-term energy neutral operation of solar energy harvesting systems

This work proposes a novel approach to dynamically adjust the system's performance level such that energy neutral operation, and thus long-term uninterrupted operation can be achieved, and achieves a two-fold improvement in system utility when compared to only applying appropriate capacity planning.

Online Human Activity Recognition using Low-Power Wearable Devices

This paper presents the first HAR framework that can perform both online training and inference, and starts with a novel technique that generates features using the fast Fourier and discrete wavelet transforms of a textile-based stretch sensor and accelerometer data.

A Survey on Human Activity Recognition using Wearable Sensors

The state of the art in HAR based on wearable sensors is surveyed and a two-level taxonomy in accordance to the learning approach and the response time is proposed.

Pack & Cap: Adaptive DVFS and thread packing under power caps

Pack & Cap is proposed, a control technique designed to make optimal DVFS and thread packing control decisions in order to maximize performance within a power budget and is implemented and validated on a real quad-core system running the PARSEC parallel benchmark suite.

Physical Activity Monitoring for Assisted Living at Home

This work proposes a methodology to determine the occurrence of falls from among other common human movements by employing classifiers based on neural networks and k-nearest neighbors and explores several effective key features that can be used for classification of physical movements.