Vayyar Imaging's new SoC differentiates between people and objects, offers high-resolution 3D imaging
LONDON — A new chip launched by Vayyar Imaging integrates an unprecedented number of transceivers and an advanced DSP to create high-resolution mmWave 3D imaging contours with high accuracy. The company claims that this breaks through current constraints in today’s 3D imaging sensor technology.
The advanced CMOS SoC covers imaging and radar bands from 3 GHz to 81 GHz with 72 transmitters and 72 receivers in one chip, enhanced by an integrated Tensilica P5 DSP with large internal memory. The company says that execution of complex imaging algorithms is all done without the need for any external CPU.
“We had to overcome several architectural challenges to achieve the overwhelming number of radio channels supported by a single chip, in terms of area, interconnects, and power consumption,” said Raviv Melamed, co-founder, CEO, and chairman of Vayyar, in an interview with EE Times.
Melamed said that the company worked closely with Cadence, TSMC, and Advanced Semiconductor Engineering (ASE) to develop the RF/analog/digital ASIC implemented on a 65-nm LP process from TSMC.
“Radio wave imaging is a powerful technology, which was dormant for decades. Vayyar’s new sensor is finally unleashing its potential,” he said. “The SoC is a multichannel radar chip supporting millimeter-wave operation as well as UWB operation. The chip has the provisions to generate signals; transmit, receive, and digitize them; and then process the received signals with an on-chip DSP processor to obtain the spatial image of the environment.”
Melamed continued, “We are bridging the gap between radar and optics. Radar has typically offered low resolution. Our chip now brings a low-cost solution to address both resolution and penetration.”
Vayyar’s sensor differentiates between objects and people, determines location while mapping large areas, and creates a 3D image of the environment. The sensor can simultaneously detect and classify a variety of targets in real time. By using wideband radio waves, the sensor can penetrate different types of materials and operates in every weather or light condition, making it applicable for automotive and industrial markets.
Melamed said, “Together with the chip, we provide our customers with a full suite of software and advanced algorithms to expedite their ability to develop products based on our technology.” He told EE Times that the first applications that they are seeing are in smart homes and in automotive.
“In the home, there’s a lot of tech and a lot of sensors. If an old person living alone falls at home, just one sensor can cover the entire home, removing the need for lots of tech. What’s even better is that it does away with the privacy issues raised when there are cameras — with radio wave 3D imaging, there are no cameras, so you can cover the home or many environments without compromising privacy. This will open up many areas for monitoring without the related security issues.”
— Nitin Dahad is a European correspondent for EE Times.