Kind: captions
Language: en
over thousands of years we've gotten
more and more efficient at growing food
for an Ever growing
population but the road from Farm to
Table can be long and
wasteful globally a third of all crops
go bad before they reach the table and
with food production accounting for
about 30% of global greenhouse gas
emissions reducing food waste could be
one solution to our climate problem at
least that's the idea behind a Norwegian
rot sniffing
robot the bomma Food Warehouse in Oslo
Norway an person is the director of
quality
assurance we get about 2,000 pallets in
here every night the produce comes in
from 80 countries they're being scanmed
here and then they go straight to the
quality control
tower this is the first control that is
being done when it comes to
Norway inspectors screen the produce for
spoilage as best they can before sending
it to the
supermarket the problem is we don't have
very much time to inspect the pallets
it's maximum 60
seconds and also due to the setup of the
quality station we're only able to
control the two upper layers
maximum that means even with experience
visual inspection only goes so far are
inevitably some spoiled produce goes
undetected and gets shipped along with
the rest of the produce all over Norway
to local
supermarkets so our question was how can
we check the whole
pallet so that's when we started to look
after new
technology the goal is increased
freshness and reduced food waste if we
can detect spoilage earlier in the value
chain we are also able to do more with
the products that we might reject we can
sort them we can give them to food
banks Bama connected with tunable a
small tech company in Oslo inventors of
an artificial nose or machine all
faction device that is already in use
monitoring the amount of greenhouse
gases emitted by container ships tyu
maduma is tun's business development
manager Bama came to us they explained
that they had this problem of
determining the quality of the fruits
and vegetables being able to do it at a
large scale and being
accurate that allow us to extend our
senses so we've done that for site we've
done that for hearing so we have
microscopes we have hearing aid but
smell is still a sense that we haven't
digitalized and that's what we're doing
Christian HOV is tunable CEO when you
take a breath you're doing a multi-gas
analyzis you're pulling in molecules and
those molecules are detected by your
nose and then it's detected by your
brain to tell you what you're smelling
the challenge for tunable was to take
their existing analyzer for emission
analysis and increase its sensitivity
without making the device too big and
cumbersome to be useful on a warehouse
floor so why use smell
our noses are sensitive detectors able
to identify a wide variety of chemicals
in the air even at low
concentrations Airborne molecules can
also potentially reveal what's hidden in
the
pallets these molecules tell a chemical
story of fruits and vegetables as they
rot but the device would have to be far
more sensitive than a human nose and
able to detect spoilage more reliably
than a human
eye produce like all living things
decays after death as microbes consume
dead cells releasing volatile organic
compounds in theory the team should be
able to tune their machine to recognize
those
molecules we knew that we could look at
complex
gases we redesigned emission
analyzer and then we started testing
Ivan YULA reer is the lead engineer on
the tunable Enos project so now I'm
going to measure fresh grapes and then
some spoiled grapes to see if our Enos
can smell the difference I'll start with
collecting a sample from the ambient air
uh as a baseline for the
measurement and the noise you can hear
now is actually the compressor pump
pulling air uh into the
analyzer so now I'm going to take a
sample from the fresh grapes to see if
there is anything present there the
probe pulls in air and then compresses
it by a factor of five which increases
the density of the sample and makes
molecules easier to
detect next infrared light shines
through the sample the light then passes
through a chip that sorts different
types of molecules based on the specific
wavelengths of light they absorb which
ultimately allows the analyzer and
accompanying sof software to reliably
detect the presence and concentration of
molecules that signal spoilage with
extreme
sensitivity the reading I got now
doesn't really show any molecules
present at all compared to Ament air
which is more or less what I would
expect from fresh
fruit so now I'm going to take a sample
for the spoiled grapes we see a clear
difference we see up to 12% absorption
at the ethanol wavelength which is a
good indication that we actually smell
the rotten grapes so uh this looks
really
promising the fumes we were able to
collect we were able to see the the kind
of the signatures the engineers then
tested different kinds of fruits and
vegetables as they decayed building up a
database of chemical
profiles we saw a tomato was different
somewhat from a
banana grapes were different from
avocado for example
and we thought well this must be
[Laughter]
interesting tour bah is the founder and
chief technology officer of tunable he's
been working with micro Electro
mechanical systems for over 30 years
tunable is a component inside our
analyzers that's the tunable filter it's
used to change the wh length for the
light so we can scan the wh length and
do spectroscopy
spectroscopy it's very much like uh
tuning a radio to find a particular
station the gases are separated in the
infrared uh Spectrum just like radio
stations and then you can basically
detect each one of them so that's where
the word tunable comes
from after extensive fine-tuning in the
lab it's time for the very first field
test in the warehouse sometimes you
can't learn about all of the variables
that will be involved in an engineered
system sitting
on a desk with a pen and paper or at a
computer screen you need to go out into
the field you need to put it in the
actual environment see how it interacts
learn from that make changes and move
forward now I'm capturing admin there
and then ready to do the measurement on
the
grapes Ivan watches the screen waiting
to see The Telltale grape
waveform but the pump just wors away
and eventually he gives
up uh I don't really know what happened
here
uh for some reason um the results wasn't
as expected the first time definitely
wasn't the charm Murphy's Law
yeah we know that it works in a
laboratory environment so the big thing
now is showing that actually
works in real life and as You' seen
there's been some
challenges we tend to think of failure
as a bad thing right that's something
that is not supposed to happen happens
but if you're doing anything new failure
is an integral part of the process and
the reason for that is because we can't
perfectly predict or understand how
things are going to work in the real
world until we try
them turns out the warehouse temperature
a chilly 41° F affected the test
result the cold part we did know that it
was cold in that area but did we take in
on account enough no we didn't we should
of course have think thought about that
but uh but that's kind of the learning
that's the
process back in the lab the tunable team
re-calibrated their chip to account for
the bomma warehouse
temperature they also adjusted the
design to include the pumps that
compress the sample increasing the
density of the gas to compensate for the
lower metabolic rate of the food in the
refrigerated
environment it will be really
interesting to see if the alterations we
have made will actually do the
difference in the
field iven is back with the latest
iteration of the
Enos further testing in the lab showed
that even with the changes the machine
needs time to adjust to the conditions
in the warehouse
[Music]
now I'll let the instrument stay here
for the night to reach a steady
temperature and then we'll do
measurements
[Music]
tomorrow now after a long cold night the
system should be ready to go
[Music]
now we see absorption of light at more
or less 9.5 10 microns which um indicate
ethanol being present this really shows
that our new chip is working in this
real
environment I uses the Enos to sample
the air from various locations on the
entire pallet stack actually we see a
spike at the ethanol absorption
wavelength so that might be something
they've taken an important step step a
successful realworld test of the newest
version of the tunable
Enos I'm not the most excited guy but
um this uh this is
exciting I expected it although you
never know that's a big
win I get very excited when technology
Works still there is work ahead to make
the technology viable and most
importantly scalable we hope that we can
make them more efficient food waste is
enormous Global problem 8% of all
greenhous gases comes from food waste so
if it can be a part of the solution it's
huge