Researchers have invented a biologically inspired localization system that integrates state-of-art sensors with neuromorphic computational maps using resistive random-access memory (RRAM)
The need for microprocessors embedded with our daily routine objects has become very essential to make our lives run smoothly. This can be achieved only if the systems can operate efficiently and reduce energy wastage accompanied by extracting crucial and compact information received from noisy and often incomplete data captured from several sensors in real-time. Hybrid memristive complementary metal-oxide semiconductor (CMOS) neuromorphic architectures provide an ideal hardware substrate for such tasks which comes with in-memory, event-driven computing capabilities. Hence, a paper published in the journal Nature Communications describes how the neuromorphic approach inspired by barn owls could efficiently provide object localization by reducing power consumption by five orders of magnitude compared with the conventional localization systems based on microcontrollers.
“Our proposed solution represents a first step in demonstrating the concept of a biologically inspired system to improve computation efficiency,” notes study senior author Dr. Elisa Vianello in a news item posted on EE Times. “It paves the way toward more complex systems that perform even more sophisticated tasks to solve real-world problems by combining information extracted from different sensors.
Researchers who were partly supported by Mem-Scales have developed a bio-inspired, event-driven object localization system that integrates advanced piezoelectric micro-machined ultrasound transducer (PMUT) sensors with a neuromorphic computational map based on RRAM. The team experimented with the proposed system consisting of RRAM-based coincidence detectors, delay-line circuits, and a full-custom ultrasound sensor. They implemented the results to calibrate the system-level simulations. These simulations were then utilized to analyze the object localization model’s angular resolution and energy efficiency. The results showed much greater energy efficiency than a microcontroller performing the same task
“The goal is, as always, to get the best power efficiency for the level of performance needed by a specific application. Further improvements in energy efficiency are certainly possible with our system,” observes Dr. Vianello.
Click here for the Published Research Paper
