Energy-Aware Duty Cycle Management for Solar-Powered IoT Devices

Abstract

IoT devices with sensors and actuators are frequently deployed in environments without access to the power grid. These devices are battery powered and might make use of energy harvesting if battery lifetime is too limited. This article focuses on automatically adapting the duty cycle frequency to the predicted available solar energy so that a continuous operation of IoT applications is guaranteed. The implementation is based on a low-cost solar control board that is integrated with the Serverless IoT Framework (SIF), which provides an event-based programming paradigm for microcontroller-based IoT devices. The paper presents a case study where the IoT device sleep time is pro-actively adapted to a predicted sequence of cloudy days to guarantee continuous operation.

Keywords: computing continuum; duty cycle; energy harvesting; internet of things; serverless.

Figure 1

The scheduler of the SIF Device Platform (SIF-D) manages events, event subscriptions, and function invocations of sensor node applications. It forwards function invocations (inv) to the Dispatcher for execution by a thread from the thread pool ones all resources are available. The Resource Manager provides the required information to the scheduler and switches resources to lower energy states when not used. The energy manager keeps track of the battery state based on the Solar Control Board. It also executes the algorithm combining solar prediction and battery state to determine the sleep time that guarantees continuous operation.

Figure 2

TUM developed this board from off-the-shelf components. It combines a three channel power IC (INA3221) (A), a battery gauge IC (MAX17048) (B), and a solar management IC (BQ25185) (C).

Figure 3

Schematics of the solar control board. (a) Schematics for the solar management chip (BQ25185). BQ24074 was replaced by BQ25185 due to chip availability. (b) Schematics for the battery gauge (MAX17048). (c) Schematics for the power measurement chip (INA3221).

Figure 4

The charging model of the 5000 mWh LIPO battery. It presents the cubic function determined by curve fitting and used as the $F_{pcharge}$ in the experiments.

Figure 5

Energy per hour of the load depending on the sleep time.

Figure 6

The energy harvested from the solar panel used in our experiments changes over the year. This diagram shows the models obtained in November 2024 and February 2025.

Figure 7

Prediction quality for each day. The graph shows the harvested energy in black, the predicted energy for the given solar panel and position on that day in red, and the prediction on the days before in greenish color.

Figure 8

These graphs present (a) the harvested energy per day, (b) the selected sleep time, and (c) the resulting RSOC at the beginning of each day for 7–20 February 2025.

Figure 8

These graphs present (a) the harvested energy per day, (b) the selected sleep time, and (c) the resulting RSOC at the beginning of each day for 7–20 February 2025.