Energy-autonomous embedded systems

“Despite the clearly visible and non-deniable global warming, it was one of these cold mornings in December 2022. Erwin H. woke up early in the dark and scrambled out of its bed, carefully not to wake up his wife. While he finally stumbled over the bedroom carpet, energy was harvested from the impact of his feet and fed into a wireless presence detector. His personal light-path to the espresso machine was turned on,and coffee production switched into operation. A water level sensor in the machine’s waste bin flashed a silent alarm to detect an upcoming overflow. His power supply was taken from a small electrochemical fuel cell in contact with the rising tide of waste water. In the shower head, a small turbine was operating to supply a wireless flow and temperature sensor. These data were directly fed into the house’s “Eco-monitor”. Erwin smiled while passing the Eco-display recognizing that he still held the family record in energy-efficient showering. On his way to the garage, he put on his wristwatch, not recognizing its internal switch-over from battery operation to thermoelectric energy harvesting from E.s wrist. The wireless car key was powered on and activated by a short “klick” on the piezoelectric generator in the “open” button. During the commuter ride ...”

Is this really science fiction?

With a closer look on our daily living and environment, we find an on-going spread-out of distributed and embedded systems, frequently microsystems, in almost every area of our daily living. More and more wireless sensors are used in logistics, environmental and building technologies, medical or automotive applications. Also, the “internet of things” is considered as the main future - and ubiquitous - network for signal processing, data transmission and control. Finally, our ever growing infrastructure networks claim for widely distributed sensor and actuator networks, e.g. for the structural integrity monitoring of bridges and tunnels or for a reliable control of our waste water management system. 

Concerning the practical applicability of such systems it is obvious that we have to equip them with energy autonomy. Due to maintenance, energy and material consumption and environmental concerns it is not conceivable to use batteries or local power grids to supply a wide-spread embedded system. Therefore, strategies have to be found for an autonomous energy supply of an embedded system from its direct environment, by tapping thermal, optical, mechanical or chemical energy. Such a “micro energy harvesting” cannot be considered as an isolated technology whose only distinction is the replacement of batteries or wire cords. In contrary, the embedded system together with its energy harvesting capabilities has to be re-designed in a full system approach as an energy-autonomous embedded system that embraces energy conversion, energy storage, energy management and the system hardware and functionality. Finally, it turns out that conventional and familiar technologies, e.g. for energy storage or power management, do not show an optimal performance in this specific application. This altogether is stimulating research and development towards optimized and application-oriented energy-autonomous embedded system, far beyond the first stage of energy harvesting.

From a more global standpoint it is interesting to see that we do not recognize embedded systems as such, however depend on their reliable operation and seam-less interaction. One might say that our natural and technical environments are enhanced via the embedding of artificial sensors and actuators, as an extension of our current natural or technical sensor and actuator systems. This vision can be carried on even further: Biological systems, the natural models of energy harvesting, generally work by a principle of "function follows energy". This way of thinking takes us to revolutionary new, biologically inspired embedded systems that live a "technical life" in their surroundings. Their design and their operating concept are made "liveable", analogous to biological principles, to maintain the function when the available energy and data vary. Like biological organisms they match their activity to the energy, use different energy sources, know their own resources, and make efficient use of them. This radical change to energy- and data-adaptive design principles promises - beyond energy autonomy - a dramatic enhancement in the operating reliability of embedded systems, and in turn opens up entirely new perspectives.