Sustainable and Green Robotics: Energy- Aware Software Architecture for Ambient- Powered Autonomous Robotic System

Most mobile and remote field robots, as well as sensor nodes, continue to operate on conventional batteries. Batteries offer flexibility, but they must eventually be replaced, which creates an e-waste problem. One approach discussed in my seminar readings is to remove batteries entirely: the robot harvests tiny amounts of energy from its surroundings, captures vibrations using triboelectric nanogenerators (TENGs), stores the electricity briefly in a capacitor, and operates in bursts. Materials for this process are developing quickly, but the software design lags behind. Embedded applications assume that power is always available and spend whatever charge has been collected in an indefinite loop that polls for charge. Here I combine research on biodegradable materials, TENGs, interrupt-driven firmware, voltage scaling, power-loss checkpointing, energy-aware scheduling, and small TinyML models that predict energy inflow. This work reports no new experiments. My aim is to put these separate ideas into one clear picture and show where they still do not fit together, chiefly the cost of the prediction model itself and the absence of proper outdoor testing. I have not built a prototype.