Energy-constrained sensor nodes can adaptively optimize their energy consumption if a continuous measurement is provided. This is of particular importance in scenarios of high dynamics such as with energy harvesting. Still, self-measuring of power consumption at reasonable cost and complexity is unavailable as a generic system service. In this article, we present ECO, a hardware-software co-design that adds autonomous energy management capabilities to a large class of low-end IoT devices. ECO consists of a highly portable hardware shield built from inexpensive commodity components and software integrated into the RIOT operating system. RIOT supports more than 200 popular microcontrollers. Leveraging this flexibility, we assembled a variety of sensor nodes to evaluate key performance properties for different device classes. An overview and comparison with related work shows how ECO fills the gap of in situ power attribution transparently for consumers and how it improves over existing solutions. We also report about two different real-world field trials, which validate our solution for long-term production use.

中文翻译：

---

感受你的力量

如果提供连续测量，能量受限的传感器节点可以自适应地优化其能量消耗。这在能量收集等高动态场景中尤为重要。尽管如此，以合理的成本和复杂性自我测量功耗仍不能作为通用系统服务。在本文中，我们介绍了 ECO，这是一种硬件-软件协同设计，可为一大类低端 IoT 设备添加自主能源管理功能。ECO 包含一个高度便携的硬件防护罩，由廉价的商品组件和集成到 RIOT 操作系统中的软件构建而成。RIOT 支持 200 多种流行的微控制器。利用这种灵活性，我们组装了各种传感器节点来评估不同设备类别的关键性能属性。概述和与相关工作的比较显示了 ECO 如何为消费者透明地填补原位功率归因的空白，以及它如何改进现有解决方案。我们还报告了两个不同的实际现场试验，这些试验验证了我们的解决方案可长期用于生产。