EDITORIAL FEATURE DIGITAL ELECTRONICS DESIGN


Words by Christian Lynn, editor of Electronics M


achines are not as advanced as people would suspect: those indulging in science fiction fantasies will claim that artificial consciousness is a present concern, when in fact, it is a distant dream. But, this is not to discourage developments in this field. In fact, recent developments in machine learning capabilities promise a new plateau for digital electronics, as they operate off of adaptable code. The digital technologies specialist, Imec, provides an example for the current generation: a compact, highly-sensitive 140GHz, multiple input/output (MIMO) radar system, functioning in support of man-machine interactions through an emphasis on gesture recognition. As machine looks to follow man, this system emulates that idea by following man’s movements, intuitively. Allowing for the detection of muscle and skin movements, Imec’s latest contribution to the digital technologies market aims to operate within vital signs applications, like non-contact patient or driver monitoring.


SMALL CHANGES In keeping with the growing trend that is miniaturisation, Imec’s radar-on-chip prototype system is notable for its small size. With a 15mm range of resolution, 10GHz of RF bandwidth, all able within an integrated, CMOS circuit, one sees a system that argues for the benefits of miniaturisation, due to the lack of performance power lost.


36 JUNE 2019 | ELECTRONICS More than this, there’s an accentuation


of the virtual within digital technology: less clunky, more liberating, is the common argument. By adding machine learning capabilities, one can recognise how the virtual begins to develop its own agency in detecting and classifying small motions within a radar system.


“As machine


looks to follow man, so does this system emulate that idea by following man’s movements”


VIRTUALLY HUMAN As activities become easier, with the rise of smart devices stripping users of a good amount of manual operation, it was only a matter of time before machines were able to learn from their routines. This is also the case with Imec’s specific machine learning algorithm, for the recognition and analysis of a number of


physical gestures. Based on a multi-layer


neural network, with


concurrent, supervised learning to train the inference model, the radar system


Imec www.imec-int.com / ELECTRONICS


In the flesh: Imec’s 140GHz radar system


bases its calculations and readings on pre-existing, labelled recordings of more than 25 personnel, made up of motion captures for seven different gestures. Delving into this history of information, Imec’s model classifies the recorded gestures and predicts the right gesture correspondingly: its success rate is noted at 94 per cent, currently. This response, based on archived footage, could be suggestive of the technology’s early stages. And yet, it still demonstrates the intelligence at work here, as the idea of memory becomes normalised in present software systems.


A VITAL FACTOR It isn’t just gestures. More importantly, vital signs can be measured, thanks to the aforementioned bandwidth. According to Imec, this subsequently opens the doors to technologies such as in-car vital sign monitoring systems, which enable the non-contact tracking of a driver’s state, such as detecting fatigue. Like a contemporary building,


constructed at an alarming speed thanks to modern technologies, digital design and technology has risen and cemented itself in terms of its capability and popularity: its inevitability is perhaps its most notable facet. Imec offers one solution: a radar system with real-world applicability, taking the digital footprint forward. As Barend van Liempd, R&D manager at Imec, concludes: “Gesture recognition can potentially lead to intuitive device control, building on an already diverse sector of existing interfaces that look to the future.”


A NICE GESTURE Digital radar systems for motion recognition and vital signs monitoring


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44