In this second part of a two-part series on the intelligence of an edge node sensor from Analog Devices, the topic of reduced time to insight is explored. As edge nodes sense information with unlimited processing power and communications data, the possibility of achieving more informed decision making seems more possible. However, because of limitations in compute cycle intensive algorithms, communications, battery power and bandwidth still persist, this possibility currently remains a concept.
Take a deeper dive into this concept as the topics of how edge intelligence is driving a smart partition paradigm shift and allowing for real-time analytics edge processing earlier in the signal chain for reduced payload burden and latency. It’s also shown how the advancements in edge technology can assist in faster time-relevant decision making as well as identifying and predicting machine performance issues. Also discussed are digital signal processing (DSP) for fixed point and floating point in regard to dynamic range, notation precision and performance, security, processing power, power consumption, duty cycle and sensor fusion.