Power profiling of autonomous industrial instruments based on virtual prototyping

Abstract

Autonomous industrial instruments such as energy harvesting temperature transmitters are slowly being adopted in industrial automation systems. These are power-critical applications, since limited amount of energy is available for the operation of an industrial instrument. Meanwhile, an instrument should be able to regularly transmit process variables during long periods of time without any maintenance. The embedded platforms used by these devices must be highly optimised in terms of low power consumption. In process industry, instruments usually consist of complex hardware and software (HW/SW). It is critical that the instrument does not break down because of the limited power supply. However, in a traditional design flow, the power consumption is hard to estimate. Virtual prototyping provides the benefits of doing simulation in the early design stage. Yet, the conventional methodologies do not estimate the power consumption efficiently and correctly. In this work, a run-time power monitoring tool is created and integrated in a virtual prototype (VP). It is based on SystemC and Transaction-Level Modelling (TLM). When the virtual prototype executes the embedded software, the power monitoring tool parallelly gives out the power estimation. The whole estimation is achieved in the early design phase of HW/SW co-simulation. The estimation result improves the HW/SW design of the instruments with optimised energy consumption. This shortens the time-to-market of the instrument, and reduces its development budget.