Page 30 - EE|Times Europe Magazine - December 2020

30 EE|Times EUROPE

Artificial Empathy in the Next Decade

and extend the human-machine relationship,
devices must become more intuitive and mas-
ter values such as caring and compassion.
“We will feel closer together with machines
that are able to receive and transmit empa-
thetic information. The end purpose is to
digitally predict the intent of the user and
allow him or her to make decisions in the
physical life based on his or her own data and
connected data.”
This is already happening in the automo-
tive domain. At the 2019 Frankfurt Motor
Show, French automotive software supplier
Valeo unveiled Smart Cocoon, which leverages
AI and the information captured by various
sensors to deliver personalized comfort
through a combination of temperature, light-
customers are asking for. “Most of the innovation that will occur for ing, sound effects, and fragrance. The interior lighting, for instance,
MEMS in the next decade will be through application-level process- adjusts automatically to reflect the temperature and adds to the feeling
ing and intelligence at the edge,” Fabrowsky said. Having AI locally, of comfort, with warm shades for heat and pale shades for cool. The
directly in the MEMS device, will yield more relevant and useful appli- vehicle becomes empathetic or emotionally intelligent in the sense
cations for users. that it considers the state of its driver and passengers by detecting
Running AI algorithms at the edge offers four main benefits. The first signs of fatigue, distraction, emotion, and stress, the company claims.
one is personalization, wherein calculation is performed locally and Building advanced driver-assistance systems “taught us that the
optimized for the user based on his or her individual behavior and envi- only way to have a real-time experience is to process the sensor data
ronment. Second is trust, or privacy of user data. Local edge processing locally,” Fabrowsky said. “All these capabilities of developing intent are
provides full control of what data is released and for what purpose. The growing in the sensor itself — not in an app, not in the cloud.” Such
third benefit is real-time feedback. Executing at the edge avoids data localization will provide better personalization, the highest privacy, the
transfer and reduces latency. Fourth is battery life improvement, enabled lowest-possible power consumption, and minimum latency.
by local processing. Germany’s Bosch Sensortec is exploring the promise of neuromor-
MEMS sensors and actuators are incorporating ever-more intel- phic sensing and computation by building brain-inspired AI at all levels
ligence, and local sensor data processing will continue to shape the of the systems, from the sensor itself to the application layers. “This
timeline, said Fabrowsky. By capturing and analyzing the wealth of helps us mimic the brain hierarchies,” said Fabrowsky.
data, sensors bring humans and the digital world together. Human- With the emergence of wearable medical devices and biosensors,
machine interactions are rapidly evolving, but isn’t it time to deepen algorithmic processes help people learn about their personal health
this relationship and make the machine identify, infer, interpret, and and fitness level. Going well beyond step counting, step tracking, and
feel human emotions? A kind of artificial empathy? activity monitoring, the next frontier is early-stage medical diagnos-
Empathetic sensors are already here, said Dimitrios Damianos, tics, said Fabrowsky. “We are just at the beginning of what empathetic
technology and market analyst in the photonics and sensing division computing can do, and new opportunities are in our reach,” he said.
at Yole Développement, in a discussion with EE Times Europe. But he These include “the ability to detect the onset of diseases like Parkin-
urged us to think about the meaning and understanding of the word. son’s or sclerosis well in advance thanks to non-invasive, always-on
“What is it to be empathetic? A human quality in the sense that it will motion sensors containing AI processing.”
understand my emotions based on my voice” and a sensing capability Eventually, artificial empathy will make devices behave “like a
combined with large processing power to read between the lines and trusted friend that we want to have at our side” and gradually enter all
decode the true state of a user’s needs, Damianos said. aspects of our lives. Applications could span across multiple sectors,
including industrial automation, robotics, telepresence, elderly care,
ACHIEVING ARTIFICIAL EMPATHY remote education, and training.
Humans are social creatures, and emotions govern our actions and From artificial strength, we have moved to AI. We are now undergo-
reactions. The ability to recognize emotional facial and vocal expres- ing a new transition from AI to artificial empathy. “In the 1950s, the
sions is innate, and when we meet someone for the first time, he or she transistor enabled the roadmap of increasing computational power,
only needs to speak for seven seconds for us to form an opinion. Read- augmenting our ability to model a forecast of the world and leading us
ing emotions accurately on the faces of others is the result of complex to AI,” Fabrowsky said. “Thanks to MEMS, we will enormously extend
processes, and neuropsychology studies suggest that intonations and and process sensory inputs, and interactions with machines will be
body language are more meaningful than the actual choice of words. more humanlike.”
Holistic perception is only possible through multi-sensor data fusion. As Fabrowsky concluded, “The next decade will be remembered for
“Sensors can independently identify and characterize an activity such its transition to artificial empathy.”
as walking, cycling, or even sleeping,” said Fabrowsky. “At the sensor artificial empathy will differ from human empathy. Could artificial
But just as artificial intelligence differs from human intelligence,
IMAGE: SHUTTERSTOCK determine the health, the feelings, and the mood of the user. Sensors empathy eliminate cognitive biases? Unlike humans, could it make
level, we are beginning to consolidate multiple information [types] to
can answer questions such as, ‘Is my stress level rising?’ and we can use
decisions devoid of cultural and gender stereotypes?
That debate, venturing into the realms of philosophy and sociology,
the information to coach the user.” This is sometimes referred to as the
tactile internet, in which all human senses interact with machines.
is one for another time. ■
“It is only by understanding such emotions that machines can
become intentional and helpful,” added Fabrowsky. But to build trust
DECEMBER 2020 | www.eetimes.eu Anne-Françoise Pelé is editor-in-chief of EE Times Europe.