What is intermittent computing?
Intermittent computing
Intermittent computing is a computing approach that leverages intermittent or unreliable power sources. Unlike traditional computing systems that rely on a continuous and reliable power supply, intermittent computing systems are designed to operate with sporadic or limited power availability. In intermittent computing, the system operates in a highly energy-constrained environment where power availability fluctuates or is intermittent. The key idea is designing hardware and software techniques to manage and leverage these intermittent power sources effectively.
Intermittent computing is needed because it allows highly low-power devices to operate solely using energy extracted from the environment. This allows such systems to operate without batteries or other external power sources. It makes them ideal for use in remote or inaccessible locations where replacing batteries is difficult or often impossible.
Use cases
While intermittent computing has a lot of use cases, let's look at some of the most common ones listed below:
Agriculture: In smart irrigation systems, intermittent-powered IoT devices monitor soil moisture levels and optimize irrigation in areas with unreliable power.
Asset tracking: IoT devices with intermittent computing track vehicles, shipping containers, and equipment. The devices can operate in diverse environments with intermittent power.
Disaster management: Intermittent-powered IoT devices collect critical data during emergencies, such as seismic activity detection or tracking rescue teams.
How does it work?
Energy-harvesting devices typically power intermittent computing systems. These systems initiate program execution once they have accumulated enough energy and halt when the energy buffer is depleted. The system's operation is divided into computation and non-volatile storage.
During this phase, the system utilizes the available energy to perform computational tasks. It carries out operations, executes algorithms, and processes data as long as sufficient power is provided. The goal is to maximize the utilization of the available energy and complete as much computation as possible before the power supply diminishes or becomes unavailable.
When the power supply becomes insufficient or intermittent, the system enters a low-power sleep state to conserve energy and preserve its current state. During this phase, all the data, intermediate results, and system state information are saved to non-volatile storage, such as flash memory. Non-volatile storage retains the information even when the power is completely cut off.
Demo
The simulator below lets us interact with a node operating on intermittent computing principles. We can toggle their states between on and off by clicking the "Node" or "Energy Harvester" buttons. Observe how the node saves and restores the work to non-volatile storage based on the battery level. Additionally, the battery charging rate decreases when the node and energy harvester are turned on.
Looking ahead
The progress in energy-harvesting technology and ultra-low-power computer systems has given rise to a new generation of IoT devices. Operating without batteries, these devices can harness energy from their environment and function intermittently. As a result, they are not bound by the limitations of batteries in terms of cost, size, and life span. This breakthrough opens up exciting possibilities for more efficient and sustainable IoT solutions.
Free Resources