Page 12 - ME-07-E

Research the field of online commerce that is about to change
(imagine yourself digitally squeezing a baguette
from the bakery and comparing its softness to a loaf
of rye). Think about computer games that allow you
to hold objects and feel their texture; TV shows that
immerse you not only in images and sound, but also
sensation; or researchers being able to “enlarge” mi-
croscopic objects such as bacteria and touch them
in order to better understand them. “What I am most
looking forward to is seeing how a system like this
can help in the medical world, whether it be for reha-
bilitation, laparoscopic surgery or training doctors,”
says Zelnik-Manor. “As far as I’m concerned, that is
the ultimate vision. But if we’re talking about a more
practical short-term realization of the technology, it’s pictures into 3D models, and Alon and Yair worked on
in gaming – giving users a multisensory experience.”
“We translate a picture into an object,
then assign it physical properties”
This fascinating collaboration was born of Wolf and
Zelnik-Manor’s academic curiosity. “Alon and I had a
talk and decided to look for things we could work on
together,” recalls Zelnik-Manor. “We had already col-
laborated in the past: For example, we co-advised a
1 2 34
HUGO, sensory feedback system. 1. Wearable version; 2. Mouse and computer tabletop version; 3. CAD model of the
tabletop version showing the system of axes in the device; 4. User interface with avatar of the user’s hand. In this
example, the interface simulates an e-commerce website.
student who researched what grabs the attention of
owls, then built an algorithm for a robot whose be-
havior is inspired by how they navigate. Alon works
in robotics and I work in computer vision, so the re-
search incorporated both fields.
“When we thought about what to do next, we
came up with digitizing touch. Alon’s specialization
is understanding how people perceive things through
touch; but touch is coupled with what we see, and
digitizing it actually requires making it possible for
people to touch things they can see virtually. In other
words, the first stage is photographing the world and
12 | MEgazine | Faculty of Mechanical Engineering
understanding what we see, which is my part; and
the next stage is translating this into a device that en-
ables touch, which is Alon’s part. We took Yair as our
PhD student for this research.”
Today, if we want to shop online for a couch, for
instance, our ability to assess its properties remotely
is relatively limited. Yes, we can ask for pictures and
a video. But what if we could sit at home and look at
the computer screen, which would give us not only vi-
sual information about the couch, but also let us feel
the texture of its upholstery? “To make this possible,
the software must have a 3D model of the couch,” ex-
plains Zelnik-Manor. “My specialization is converting
developing the touch devices. In research, it’s easiest
to study each component separately at the highest
possible level – and only later, at a more advanced
stage, bring the components together.
How do we even start planning a device that imi-
tates touch? “Our skin has receptors that are sensi-
tive to pressure, vibrations, shearing forces and tem-
perature,” explains Prof. Wolf. “Our fingertips have
an especially high concentration of these receptors.
Lihi was in charge of teaching the software the prop-
erties of various surfaces, and we were in charge of
developing a system that conveys sensation based
on the information from the picture.”
Touch, warns Wolf, is a very ‘relative’ sense: “Its
resolution is much coarser and more subjective than
that of sight, for instance.” Therefore, he explains,
expectations must be set accordingly. “If someone
uses the device to try out four different comforters,
the system still cannot provide exactly the same sen-
sation that actually touching them would – but it can
provide a precise ranking of softness, for example, so
the shopper can tell which blanket is softest to the
touch.”