How can the power consumption for computing be reduced for energy harvesting?

At Sensors Expo 2018, Chet Jewan, vice-president of sales and business development from Eta Compute, explains how his company’s computing technology can benefit heart rate monitoring and energy harvesting applications. The first demo, with an Eta Compute chip running at 0.476V powered by solar energy, displays Jewan’s heart rate.

A second demo has sensors being power strictly by solar energy without batteries. The demo wirelessly transfers light and temperature data periodically to the display. In both instances, the power consumption required for computing in energy harvesting applications is achieved by asynchronous processing of all digital logic. The efficient event-driven computing technique consumes power only when spiking. Using spiking neural network (SNN) algorithms for speech, video, sensor fusion and other applications, the company’s low power system on chip and microcontroller intellectual property (IP) brings machine intelligence to energy constrained systems.

The company’s ECM3531 is an application specific integrated circuit (ASIC) for machine learning algorithms based on the ARM Cortex-M3 and NXP Coolflux DSP processors. The chip includes a low-power analog to digital converter (ADC) sensor interface and highly efficient power management IC circuits.

For motion and environmental sensors that use accelerometers and gyros and/or a variety of chemical sensors, the company’s EtaCore technique for sensor fusion enables sensors hubs to perform more extensive sensor algorithms providing real time data and updates from mobile and IoT networked devices. The output can be connected to systems using LoRaWan, Z-Wave, EnOcean, Wi-SUN and Bluetooth standards.