As the impacts of climate change continue to escalate, public and private sector organizations find themselves in urgent need of innovation. Emergency management has been grappling with outdated tools and insufficient tech investment for decades. Companies such as Pano AI are stepping up to mitigate and better prepare for further crises with a new generation of monitoring, alert, and data analysis tools to help battle fires and other emergencies.
In this episode of The Restless Ones, I had the chance to speak with Sonia Kastner, founder, and CEO of Pano AI, who is bringing the field of disaster management into the connected age. Pano’s focus is on hardware that utilizes available technologies such as AI, Computer Vision, and 5G connectivity to create an end-to-end real-time solution for both private and public entities fighting fires in their domain.
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:03):
Hey, everyone, Welcome to the Restless Ones. I'm Jonathan Strickland.
As always, my focus is on exploring the intersection of
technology and business by having conversations with the most forward
thinking leaders. Throughout my career, I've covered everything from massive
parallel processing to advanced robotics, but what truly inspires me
(00:24):
are the stories of innovation and transformation. Today, you're going
to hear a great conversation I had with Sonia Kassner,
founder and CEO of Panoai. Her company brings together technologies
such as imaging, cloud computing, AI, and yes connectivity to
provide a very important service early detection of forest fires.
(00:49):
As you'll hear, Panoai partners with customers to deliver data
needed to identify and respond to forest fires. But before
I could geek out about Panoai's approach and technology, I
wanted to learn more about Sonia's journey. Sonia, it is
(01:10):
a huge pleasure to have you on the show. Welcome
to the Restless Ones.
Speaker 2 (01:15):
Thank you so much for having me.
Speaker 1 (01:16):
I'm really excited to have a conversation with you, So
can you kind of give me sort of a story
of your journey up to now really hitting the highlights.
Speaker 3 (01:26):
Absolutely, I've always worked at the intersection of business, technology,
and government. Early in my career I was a young
activist working on political campaigns. I also worked in the
New York City government under Michael Bloomberg. And I love
the impact that you can have in the public sector,
but I also love the innovation and speed and nimbleness
(01:47):
you can have in the private sector. And in two
thousand and seven, I decided to move out to California
and get involved in the green energy industry, which was
a booming sector in venture capital back then. I went
to business school at Stanford, I joined one of the
many venture backed solar startups at the time, and I
really hope to spend my entire career working on technologies
(02:09):
that would prevent climate change. And sadly, fast forward fifteen years,
I think everyone would agree we have not prevented climate change.
Climate change is now very much clearly.
Speaker 2 (02:19):
All around us.
Speaker 3 (02:20):
We're living it every day, and natural disasters are one
of the ways we're feeling the effects of climate change first.
And these natural disasters people are starting to rever to
them as climate disasters. Floods, hurricanes, wildfires, mud slides, extreme heat,
extreme cold, tornadoes. They're really serving as a wake up
call to humanity that climate change is here. Since then,
(02:42):
I've been working in supply chain manufacturing at companies like Nest,
where I actually worked on the Nestcam in twenty fourteen,
one of the first AI cameras. In twenty eighteen twenty nineteen,
I was thinking about ways that I could apply the
technology that I was familiar with, coming from Silicon Valley,
coming from Internet things, to help with tackling natural disasters
(03:03):
and see if we could make these natural disasters less harmful.
And as I started to do research, the good news
I found was actually there was a lot of low
hanging fruit. The natural disaster industry had not seen a
lot of innovation for decades, and there's tremendous hunger from
emergency managers such as fire agencies, such as power utility
emergency managers. They were asking send us cameras, send us drone,
(03:28):
send US satellite technology, artificial intelligence, send us better software tools,
help us use mobile phones better, because we need technology
as a force multiplier to help us tackle this growing threat.
And we looked around, there were very few companies that
have been founded out of Silicon Valley, building tools for
this industry, and we decided to answer the call, and
(03:48):
that was the story behind pano.
Speaker 1 (03:49):
Wow and speaking of it another things. Since I have
you on here, I am somewhat curious as to what
your first exposure to that concept was, because it's rare
that I talk someone who is so deeply involved in
the Internet of Things.
Speaker 3 (04:04):
Absolutely so I got involved in the Internet of Things
industry early on. I had been working in clean tech,
which went through a boombust cycle, and found myself needing
to find another industry to work in, and I started
hearing about Internet of Things, how Internet connectivity like Bluetooth
and Wi Fi and cellular technologies were going to move
(04:26):
from just being in your phone and computer into other
types of devices and really IoT. The theme of IoT
has always been gathering data that could be used for AI.
IoT and AI have actually always been coupled. My first
IoT job was at a company called Whistle, where we
were making a fitbit for dogs, and we thought it
(04:47):
was going to be an easy product to make because
there was already a fitbit for humans. But it turns
out a fitbit for dogs came with its own new challenges.
For example, dogs don't carry phones, so Fitbit had Bluetooth
to just the data back to your phone, but your
dog was home without a phone, so we actually had
to put Wi Fi into this dog collar device, and
(05:08):
in hindsight, it turned out to be one of the
first IoT devices that had miniaturized Wi Fi. I think
we were one of the first users of the Athos
Qualcom chip and it was hard, very very hard. But
we shipped it in fifteen months and it was a
great introduction to the world of IoT and that was
where I spent my career for the next ten years.
Speaker 1 (05:28):
Excellent. I love that story. So you know, I'm curious
because before you founded Pano, you were an angel investor
and you were specifically focusing on startups that were centered
around hardware, which to me is mind blowing because I
think anyone who's ever been in tech knows hardware is hard.
It is hard to get that right, it's a long
(05:48):
development cycle. What kind of drove you in that direction
to look for companies that you could help establish and
support that were in the hardware space.
Speaker 3 (05:58):
I love working in hard and that's what we do
here at Pano as well. There are certain things that
make hardware harder, but if you know how to do them.
Then it becomes a great moat, and I think every
startup needs to find a moat. I think there's a
great article from one of the y common Air founders
that every startup needs its schlep right, and if there's
no schlep, then there's gonna be too much competition, it's
(06:20):
not going to be a good business. You just have
to pick what your schlep is going to be. So
the schlep of hardware is a couple of things. First
of all, in a hardware company, much more of the
work you need to do is done with third parties
for other companies, not within the own walls of your company,
not by employees. So your engineers are mainly going out
(06:40):
and selecting technology from third party companies, and they have
to then work with those vendors, they have to work
with contract manufacturers. So it's a team sport outside of
your company. It's also much more of a team sport
within the company. There's much more interdependencies between mechanical engineers
and electrical engineers then necessarily between discrete software engineers, so
(07:01):
that makes it challenging. It's much harder to change scope
of a hardware product once you've started, so you have
lex fusibility there, and you also have more specialists within
your team, so the people on your team are not fungible,
so you've fixed your scope. You often have a hard deadline,
like you're shipping for Christmas. In our case of pano
we're shipping for fire season, and things go wrong a
lot in hardware and you have to sprint and recover.
(07:25):
I think it tends to be a good field for
people who are a bit of adrenaline junkies, where you know,
you make really really careful plans and then everything still
goes wrong and you have to figure out how to react.
My thinking is that actually is pretty similar to disaster management,
and that's actually part of also what drew me into
getting into the disaster management industry as well.
Speaker 1 (07:44):
Wow, great insight, And we've already alluded to the fact
that pano is in hardware space. That panel relies a
lot on hardware. Can you talk a bit about the
types of hardware PANOAI is using in order to detect
forest fires. What actually is the technology that's being used
out in the field.
Speaker 3 (08:04):
Yeah, I can describe the hardware and maybe how the
whole system works as well. So Pano's solution is a
solution for detecting and alerting on new fire ignitions within
minutes and pushing out alerts that come with time lapse
video of the fire growing and location information GPS coordinates
(08:24):
so that first responders can see the fire, make an
assessment about what kind of response is necessary, share the
fire information with each other over text and email, and
then create a coordinated aggressive response to send helicopters, planes,
bulldozers go nip that fire in the butt is actually
modeled after an scam or a ring doorbell, very very
similar idea. From a user's perspective, it's almost the same
(08:47):
as a nestcam or ring doorbell that says, somebody's stealing
your package.
Speaker 2 (08:50):
Here's a video.
Speaker 3 (08:51):
And actually, I think it's a big part of why
the product's been so popular and being widely adopted because
the customers actually are used to this concept. To the
customer seems really simple. Behind the scenes, the whole system
is incredibly complicated. So it starts with a piece of
equipment that we design a manufacturer called a Pane station.
This includes about forty components, including two high definition security cameras.
(09:15):
These are off the shelf top of the line six
megapixel security cameras designed for ruggedized environments, and about forty
other components. We assemble them in our factory in San Francisco.
The systems also include an edge computer which has logic
on how we control the cameras. We also include networking
equipment so we can send the data up to the
cloud over cellular broadband connectivity. We have power management. Sometimes
(09:40):
we need to include a backup battery. Sometimes we need
to include solar panels. Every site is a little bit different,
so it's a configured to order supply chain, which is
not for the faint of heart.
Speaker 2 (09:50):
And we mount.
Speaker 3 (09:51):
These systems typically on existing structures like cell towers, water tanks,
government communications towers, sometimes even private home homes or chairlifts
at a ski resort. We get really creative on where
to put these and we've actually found five G to
be a really great technology for us. One of our
investors is t Mobile, and they've been a phenomenal partner
(10:12):
in terms of allowing us to migrate from four G
to five G, and we're one of the first industrial
users of five G technology, which has further range into
the forest and higher performance than four G. So Once
we mount these systems, we rotate the cameras three hundred
and sixty degrees every minute, and we take ten frames
with each rotation, and we send this up to the cloud.
(10:35):
We put these images through a computer vision deep learning
algorithm and it draws bounding boxes on the images where
it thinks there's smoke. Each one of these bounding boxes
is reviewed by a human analyst in our Paneo Intelligence
center to make sure that it's really truly smoke. This
is called human in the loop AI. It's actually a
(10:55):
very common approach with computer vision AI because it's so
easy for a human to use damage as ground truth
to be sure.
Speaker 2 (11:02):
In that way, when.
Speaker 3 (11:03):
We trigger an alert out to our customer, we have
almost perfect signal to noise ratio, so the customers know
when they get an alert from PANO, this is going
to be smoke, and I don't have to worry about
being spammed. Then the customers get all the other rich
information they need to go control the fire.
Speaker 1 (11:18):
I'm glad you brought up the human loop element because
anyone who's followed any sort of AI, even things like
image recognition, you realize that there are outlier cases where
I think of it as like Paradelia where you're looking
up at the sky and you see a cloud and
it is in the shape of something, and as a person,
you know that's not actually the thing. It's not really
(11:40):
a cat up in the sky. It just looks like one,
but to an image recognition algorithm, it might fool it
into thinking it's a cat. And so having that human
element to be that step before sending messages out, before
you start to see agencies commit resources toward going to
a potentially remote location that is not easy to get to,
(12:00):
which could be taking their attention away from other very real,
very present emergencies.
Speaker 2 (12:06):
Yeah.
Speaker 3 (12:06):
Absolutely, And the good news is cameras as an AI
sensor are really phenomenal. And that's actually a core part
of Pano's long term vision and strategy is we've decided
to specialize in computer vision AI. We just hired an
amazing VP of engineering. He has a PhD in computer vision,
and we brought on a leader with computer vision and
(12:28):
expertise to lead our engineering effort because we think that
camera sensors are going to be the most important for
emergency management, whether those cameras are stationary like our first product,
or whether those cameras are on a drone or satellites.
With camera data, you can run your algorithm and you
do object detection to detect the signal you're trying to detect,
(12:50):
but you can also have a human verify in real time.
Because the camera data is ground truth, you can combine
artificial intelligence and human intelligence together in real time with
camera data beautifully, and so that makes it, I think,
much more useful and much more powerful.
Speaker 1 (13:12):
You also mentioned five G connectivity, which is obviously something
we really like to talk about on this show, and
it's interesting because you were bringing up a different sort
of flavor of five G than what I often talk about.
I'm often talking about the ultra high frequency five G,
where you have very high throughput, very low latency. But
for those who do not know, there are other bands
(13:34):
of five G, and as you were eluding, one of
those bands is one where you are able to get
much further broadcast range than you would with other cellular technologies.
So you mentioned about the partnership with T Mobile. Can
you talk a little bit more about that and how
five G has played a part in pano AI's operation.
Speaker 3 (13:54):
Yes, absolutely, so, we've been partnering with T Mobile for
over a year now. I think we were in the
twenty twenty two class of the five G Open Innovation
Lab and they help facilitate a partnership with T Mobile
over a year ago, and that has been a really
successful partnership. T Mobile helped us with technical support that
allowed us to migrate from four G to five G
(14:17):
sooner than almost any other industrial application, and this has
really been terrific for us, and I think it's just
going to continue to be more powerful over time as
we unlock more and more features that take advantage of
the performance of five G. So the benefits for our
system of five G are, first of all, the range
as you referred to. We're installing our PANA stations on
(14:40):
mountaintop locations, typically at the wildland urban interface or deep
in the forest. That's where you need to have wildfire cameras.
You don't need to have wildfare cameras right and downtown.
And sometimes our locations are not very close to a
cell tower. They can be far from a cell tower.
We could be mounting on an abandoned four service lookout
tower that's miles into the forest. So the fact that
(15:00):
five G provides for further range from the cell tower
means that we might be able to use cellular in
a location where otherwise cellular would not be an option.
Second of all, the latency and the broadcast and the
upload speeds and the bandwidth are actually very beneficial for
us as well. I mean, remember we're an upload application,
not a download application in terms of latency. This really
(15:21):
comes into play when we're using the cameras in optical
zoom mode. So our standard operating mode is patrol mode.
We're rotating the cameras one rotation per minute, scanning the
landscape looking for smoke. But once we find smoke, then
either Pano or our customers will take one of those
two cameras out of rotation and zoom in on the
(15:44):
smoke using the very powerful optical zoom capabilities of the camera.
And this actually provides much richer visual information of what's
going on with the fire. It allows the fire agencies
to much more clearly see road access, to see which
way the wind is blowing, to see strokes nearby. It
can really have a huge impact on firefighter safety. And
(16:05):
the challenge is as you're going to navigate the camera,
if you're on a four G connection, you have to
wait to send the instructions out to the camera, and
it can be a nuisance and slow down the navigation
and take you longer to find the fire and zoom
in to where you want to get to. It's not
the same as performing open heart surgery, but the latency
really will be of great advantage. And then finally, in
terms of the upload bandwidth, these cameras are capable of
(16:28):
thirty frame per second video, but on four G we're
typically uploading about ten frames a minute. Depending on the
four G connection we can go a little bit above that,
but with five G we can go to the full
thirty frame per second video, which, again, when you're staring
at a fire, you want all the visual information you
can get.
Speaker 1 (16:45):
Yeah, that's incredible. Well, can you tell me a little
bit more about the technologies that are on the back end.
I'm assuming that you have a mixture of cloud computing
and on premises compute that are processing all this images.
Can you kind of give us a big picture of
how that works.
Speaker 2 (17:00):
Sure.
Speaker 3 (17:00):
I love the theme that a technology is developed by
one industry and then other sectors find ways to harvest
that technology and use it in new ways. So at
PANO we're harnessing many technologies that were developed for other industries.
Five G is a perfect example, was developed primarily for
cell phone users, but we're using it for wildfire detection.
(17:22):
The AI that we use was really primarily developed for
the self driving car industry, and we're harnessing that for
fire detection. Satellite technology is really critical for us. We're
using geostationary satellite data that was developed by Noah, the
National Weather Service. Those have been very useful as a
supplemental detection method using thermal sensors. We're harnessing cloud tools.
(17:46):
I mean we ingest massive amounts of data. If we
were not able to harness scalable cloud technologies and Kubernetes,
there's no way we could do what we're doing today.
We use SaaS tools like Twilio to push out our
notification and scalable fashion. So there's dozens of modern day
technologies that were developed over the past ten years that
(18:07):
are included in our solution that our customers are totally
unaware of don't have to think about as Our job
is to go shop for the best technologies like five
G and bring them to our customers as as soon
as they become readily available.
Speaker 1 (18:19):
I love that explanation as well. It is another inspiring
approach to this and the idea of using that specifically
for a climate oriented task here where we know for
a fact that we're going to have more of these
climate events. It's interesting to me about your business because
when I first heard about what PANO does, to me,
(18:41):
it was immediately one of those things where I thought, Oh,
that sounds like something that you would think of as
an adjunct to a government agency for example. Can you
talk a bit about what thinking went into making it
a business, how did that sort of blossom?
Speaker 3 (18:57):
So it absolutely is a service that should be provided
by government agencies, so you're absolutely right there. And then
government agencies typically select vendors by running an RFP, and
that's how they select the vendors that then deliver the
government services that you're used to. And actually we've been
very impressed with the rigor that government agencies have used
(19:17):
when selecting vendors.
Speaker 1 (19:20):
Excellent. Can you talk a bit about sort of the
skills that are needed for a business leader who wants
to work closely with governments? What sort of skill set
does that person need and what sort of expectations should
they have?
Speaker 3 (19:33):
Yeah, so I had worked in government, but selling into
government was not my skill set. The first person who
joined our team, I'm very fortunate, is our chief commercial officer,
Arvinsattiam and he did have ten years of experience selling
into government. He had been working at Cisco for many years,
and for the past ten of those years he was
one of the leaders of their government IoT business, where
(19:56):
he sold million dollar deals into sy like the city
of Barcelona that led to wiring up the entire city
to be a smart city with smart street lamps and
smart parking meter, smart trash cans, and so he is
exceptional at learning how to navigate the challenges of selling
into government buyers. So I would say that selling into
(20:17):
government we have found to be more difficult than selling
into any other segment. So you need both sellers and
you also need regulatory We have regulatory advisors as well.
But I will also say a really important part of
Pano's business success has been that we don't only sell
to government. So our business model is that we deploy
(20:37):
in a region and we sell subscriptions to multiple types
of entities. So we sell to government agencies city, county, state, federal,
but we also sell to power utilities. They're one of
our largest customer segments. They have a lot of wildfire risk.
Their power lines run through high fire risk areas and
so they're really focused on protecting their ass that's we
(21:00):
also sell to other property owners like ski resorts, the
forestry industry, other private landowners, real estate developers. So it's
been really important as a venture backed company that we
have a mixture of public sector and private sector customers
in order to meet the business objectives. And also actually
(21:21):
we found these two types of buyers to be complementary.
We found that our private sector customers appreciate the fact
that we also have a public sector business. In fact,
we've done multiple deals where we've actually signed up both
a public sector and a private sector customer at the
same time. For example, in the Telluride area, we signed
up the Telleride Fire Department and SA Miguel Power Company
(21:44):
as well at the same time for the pilot. This
is called a public private partnership and both parties feel
great about this collaboration, and it's because the fire agency
feels great that the utility is participating and contributing to
support protecting their assets, and the power utilities know that
if the fire agencies are participating as customers that they're
(22:07):
going to use this product because the enterprise customers can't
suppress the fire on their own. The end users of
the product are the fire agencies.
Speaker 1 (22:14):
It's really interesting. It also makes me think of cases
where there are issues with power utilities and their close
association with fires. I'm reminded of rolling brownouts that were
needed in times of high winds, for example, where there
was a concern that utility lines could potentially cause fires,
and potentially down the line, a company like PANO could
(22:37):
end up delivering lots of information that could lead to
things like infrastructure overhaul. Like it's interesting to me how
your role could expand beyond the immediate detection of fires
but also ultimately contribute to improving infrastructure in places that
need it. And it's because you're able to gather the
information that's necessary before anyone can make those kinds of decisions.
Speaker 3 (23:01):
Yes, you're spot on. I mean, once again, I think
you know our long term vision better than we do exactly.
You know, our first solution at PANO is focused on
delivering real time, actionable intelligence during the response phase of
an incident. And by the way, I should also mention
our customers use our solution during fire season for enabling
(23:22):
rapid suppression of fires, but they also use our solution
in the rainy season for enabling safer and more manageable
controlled burning. In fire season, our solution is used for
pushing out real time actionable information to help with suppression,
and that's called the response phase of a natural disaster.
But there's three other phases of natural disaster management. There's
(23:44):
the recovery phase. There's the mitigation phase, which is hardening
your system, which would be for example, decisions around bearing
power lines or relocating assets, for example, making a decision
that you're going to change your deck material to make
it non flammable.
Speaker 2 (23:59):
That would all be part of mitigation.
Speaker 3 (24:01):
And then there's preparedness, which is okay, making sure you
have your evacuation routes planned out, making sure you have
a way to contact people, making sure you have shelters
with enough supplies. That's all preparedness, and then all of
that feeds back into the response and the cycle continues.
To your point, the data that we are creating we
create in our tool. After we push out an incident,
(24:22):
that data doesn't just go away. We've created an archive
of all of these fire incidents that we've alerted on,
so we have a very accurate record of exactly where
and when the fire started, and we ingest third party
data as well that allows us to have data on
how that fire was responded to. We ingest all of
that into our database, and we now have a treasure
(24:45):
trove of fire activity fire response all over the country.
And as our deployments increase, that data is just growing
and growing. And when I think about the mission of
PANO and the impact that we can have, I see
as twofold. One impact we can have is the impact
of improving real time response and lessening the harm of
wildfires in the immediate term, but I believe that we
(25:06):
can have additional impact by improving hardening decisions. There's going
to be trillions of dollars of investment made to harden
our infrastructure and rebuild our cities, rebuild our transportation in
the face of climate change. It's going to cost billions
of dollars to bury power lines to create microgrids. That's
(25:27):
going to take decades, and we're going to need to
spend billions for new helicopters and new planes. How are
we going to decide how to prioritize those investments, and
which investments are needed and which ones are not needed.
All of that needs data. There's no data today. There's
a federal agency that created a national Fire Incident Database,
and they created a database with ninety six fields that
(25:49):
should be populated for every fire incident. And when you
look at the database, only two of the fields are populated.
One field is when was the fire called in and
how many acres was it when it was And even
those two fields are suspect, right, how do we know
if that's really accurate information? But with Panos technology, you
can absolutely get that information accurately and many, many more
(26:11):
of those ninety six fields can be populated as well,
and data can really be an enabler for us to
adapt to climate change.
Speaker 1 (26:18):
Sonya, You've talked about so much here with the Internet
of things, artificial intelligence, the era of big data, the
ability for us to be able to do something meaningful
with that data. I think a lot of these things
all started to converge in a really exciting way around
the same time. The future you're describing as one that
I expect we're going to see lots of other businesses
(26:38):
focus on. I hope to see more organizations take that
approach of how can we leverage the massive amounts of
information we collect in ways that have a real positive
outcome in the real world.
Speaker 2 (26:54):
Absolutely, you're spot on.
Speaker 3 (26:55):
I think there's a hopeful message here, which is that
we as humanity have been work working on trying to
find solutions to slow down or reverse climate change for decades,
and I think many of those I'm optimistic about, and
I think human innovation will find a way to reverse
climate change, but it's going to take decades. There has
been almost no innovation that has gone into adaptation to
(27:17):
climate change, and as a result, what that means is
that there's so much low hanging fruit. We have so
many technology tools that are perfectly suited to help cope
with the effects of climate change, sensors, drones, satellites, five G,
big data software tools, and the entrepreneurship wave of the
(27:38):
climate adaptation field is just beginning. Panel is one of
the first companies in this field, and we're hoping to
blaze a trail here and we're hoping other entrepreneurs will
see that technology really can make climate change less harmful
and that we can save lives we can save homes,
we can save trees, we can save animal life. Let's
put our heads together and think about ways that we
(27:59):
can innovate to lessen the harms of a climate change.
Speaker 2 (28:02):
It's already here today.
Speaker 1 (28:05):
I had a couple more questions to ask Sonia before
I could let her go. Well, next up, what is
the best piece of advice you have ever received?
Speaker 2 (28:19):
A professor of mine?
Speaker 3 (28:20):
In college senior year, I was taking a behavioral economics class,
and he was explaining that people do not do a
good job of intuitively understanding probabilities. So people often are
afraid of losing something, even if it's a small amount
or a low risk of losing. People are grasped what
they have. They don't want to lose something. They buy
(28:41):
lottery tickets because they think they have a chance of
winning something big. So what he encouraged us is take small,
positive expected value risks in life. And I think that
was really good advice.
Speaker 1 (28:53):
I think so too. Yeah, I like that a lot.
Generally speaking, humans are really bad at assessing risk unless
we put a lot of effort toward it, I find. Yeah,
what does the term restless ones mean to you?
Speaker 3 (29:07):
You know? I love the name of this podcast, The
Restless Ones, and it made me think of not being
satisfied with accepting the world for how it is, but
saying I don't like how the world is today. I
think the world should be different, and I'm going to
go do something about it.
Speaker 1 (29:23):
Excellent. I couldn't agree more. Thanks again to Sonia Castner
of pano Ai for talking with us on The Restless Ones.
Sonia's story reminded me that innovation isn't restricted to inventing
the next light bulb. Innovation can involve finding a new
(29:43):
way to use that light bulb to tackle tough problems.
Taking advantage of advancements in technologies ranging from IoT deployments,
image capture and artificial intelligence, big data analysis, and yes,
advanced wireless technologies makes pano Ai a true pioneer. Moreover,
I think Sonya's story reminds us that in the face
of huge challenges, there is opportunity, and if we can
(30:07):
look at climate change as an opportunity to dedicate engineering
and innovation to protecting ourselves and future generations and with
any luck, eventually reverse some of those changes, that's truly inspiring. Again,
thank you Sonia Kastner, and thank you listeners for checking
(30:29):
out the Restless Ones, be sure to look through our
back catalog of conversations with leaders at the intersection of
business and tech, and come back for more great discussions
with the future Restless Ones. I'm Jonathan Strickland.
Hey, everyone, Welcome to the Restless Ones. I'm Jonathan Strickland.
As always, my focus is on exploring the intersection of
technology and business by having conversations with the most forward
thinking leaders. Throughout my career, I've covered everything from massive
parallel processing to advanced robotics, but what truly inspires me
(00:24):
are the stories of innovation and transformation. Today, you're going
to hear a great conversation I had with Sonia Kassner,
founder and CEO of Panoai. Her company brings together technologies
such as imaging, cloud computing, AI, and yes connectivity to
provide a very important service early detection of forest fires.
(00:49):
As you'll hear, Panoai partners with customers to deliver data
needed to identify and respond to forest fires. But before
I could geek out about Panoai's approach and technology, I
wanted to learn more about Sonia's journey. Sonia, it is
(01:10):
a huge pleasure to have you on the show. Welcome
to the Restless Ones.
Speaker 2 (01:15):
Thank you so much for having me.
Speaker 1 (01:16):
I'm really excited to have a conversation with you, So
can you kind of give me sort of a story
of your journey up to now really hitting the highlights.
Speaker 3 (01:26):
Absolutely, I've always worked at the intersection of business, technology,
and government. Early in my career I was a young
activist working on political campaigns. I also worked in the
New York City government under Michael Bloomberg. And I love
the impact that you can have in the public sector,
but I also love the innovation and speed and nimbleness
(01:47):
you can have in the private sector. And in two
thousand and seven, I decided to move out to California
and get involved in the green energy industry, which was
a booming sector in venture capital back then. I went
to business school at Stanford, I joined one of the
many venture backed solar startups at the time, and I
really hope to spend my entire career working on technologies
(02:09):
that would prevent climate change. And sadly, fast forward fifteen years,
I think everyone would agree we have not prevented climate change.
Climate change is now very much clearly.
Speaker 2 (02:19):
All around us.
Speaker 3 (02:20):
We're living it every day, and natural disasters are one
of the ways we're feeling the effects of climate change first.
And these natural disasters people are starting to rever to
them as climate disasters. Floods, hurricanes, wildfires, mud slides, extreme heat,
extreme cold, tornadoes. They're really serving as a wake up
call to humanity that climate change is here. Since then,
(02:42):
I've been working in supply chain manufacturing at companies like Nest,
where I actually worked on the Nestcam in twenty fourteen,
one of the first AI cameras. In twenty eighteen twenty nineteen,
I was thinking about ways that I could apply the
technology that I was familiar with, coming from Silicon Valley,
coming from Internet things, to help with tackling natural disasters
(03:03):
and see if we could make these natural disasters less harmful.
And as I started to do research, the good news
I found was actually there was a lot of low
hanging fruit. The natural disaster industry had not seen a
lot of innovation for decades, and there's tremendous hunger from
emergency managers such as fire agencies, such as power utility
emergency managers. They were asking send us cameras, send us drone,
(03:28):
send US satellite technology, artificial intelligence, send us better software tools,
help us use mobile phones better, because we need technology
as a force multiplier to help us tackle this growing threat.
And we looked around, there were very few companies that
have been founded out of Silicon Valley, building tools for
this industry, and we decided to answer the call, and
(03:48):
that was the story behind pano.
Speaker 1 (03:49):
Wow and speaking of it another things. Since I have
you on here, I am somewhat curious as to what
your first exposure to that concept was, because it's rare
that I talk someone who is so deeply involved in
the Internet of Things.
Speaker 3 (04:04):
Absolutely so I got involved in the Internet of Things
industry early on. I had been working in clean tech,
which went through a boombust cycle, and found myself needing
to find another industry to work in, and I started
hearing about Internet of Things, how Internet connectivity like Bluetooth
and Wi Fi and cellular technologies were going to move
(04:26):
from just being in your phone and computer into other
types of devices and really IoT. The theme of IoT
has always been gathering data that could be used for AI.
IoT and AI have actually always been coupled. My first
IoT job was at a company called Whistle, where we
were making a fitbit for dogs, and we thought it
(04:47):
was going to be an easy product to make because
there was already a fitbit for humans. But it turns
out a fitbit for dogs came with its own new challenges.
For example, dogs don't carry phones, so Fitbit had Bluetooth
to just the data back to your phone, but your
dog was home without a phone, so we actually had
to put Wi Fi into this dog collar device, and
(05:08):
in hindsight, it turned out to be one of the
first IoT devices that had miniaturized Wi Fi. I think
we were one of the first users of the Athos
Qualcom chip and it was hard, very very hard. But
we shipped it in fifteen months and it was a
great introduction to the world of IoT and that was
where I spent my career for the next ten years.
Speaker 1 (05:28):
Excellent. I love that story. So you know, I'm curious
because before you founded Pano, you were an angel investor
and you were specifically focusing on startups that were centered
around hardware, which to me is mind blowing because I
think anyone who's ever been in tech knows hardware is hard.
It is hard to get that right, it's a long
(05:48):
development cycle. What kind of drove you in that direction
to look for companies that you could help establish and
support that were in the hardware space.
Speaker 3 (05:58):
I love working in hard and that's what we do
here at Pano as well. There are certain things that
make hardware harder, but if you know how to do them.
Then it becomes a great moat, and I think every
startup needs to find a moat. I think there's a
great article from one of the y common Air founders
that every startup needs its schlep right, and if there's
no schlep, then there's gonna be too much competition, it's
(06:20):
not going to be a good business. You just have
to pick what your schlep is going to be. So
the schlep of hardware is a couple of things. First
of all, in a hardware company, much more of the
work you need to do is done with third parties
for other companies, not within the own walls of your company,
not by employees. So your engineers are mainly going out
(06:40):
and selecting technology from third party companies, and they have
to then work with those vendors, they have to work
with contract manufacturers. So it's a team sport outside of
your company. It's also much more of a team sport
within the company. There's much more interdependencies between mechanical engineers
and electrical engineers then necessarily between discrete software engineers, so
(07:01):
that makes it challenging. It's much harder to change scope
of a hardware product once you've started, so you have
lex fusibility there, and you also have more specialists within
your team, so the people on your team are not fungible,
so you've fixed your scope. You often have a hard deadline,
like you're shipping for Christmas. In our case of pano
we're shipping for fire season, and things go wrong a
lot in hardware and you have to sprint and recover.
(07:25):
I think it tends to be a good field for
people who are a bit of adrenaline junkies, where you know,
you make really really careful plans and then everything still
goes wrong and you have to figure out how to react.
My thinking is that actually is pretty similar to disaster management,
and that's actually part of also what drew me into
getting into the disaster management industry as well.
Speaker 1 (07:44):
Wow, great insight, And we've already alluded to the fact
that pano is in hardware space. That panel relies a
lot on hardware. Can you talk a bit about the
types of hardware PANOAI is using in order to detect
forest fires. What actually is the technology that's being used
out in the field.
Speaker 3 (08:04):
Yeah, I can describe the hardware and maybe how the
whole system works as well. So Pano's solution is a
solution for detecting and alerting on new fire ignitions within
minutes and pushing out alerts that come with time lapse
video of the fire growing and location information GPS coordinates
(08:24):
so that first responders can see the fire, make an
assessment about what kind of response is necessary, share the
fire information with each other over text and email, and
then create a coordinated aggressive response to send helicopters, planes,
bulldozers go nip that fire in the butt is actually
modeled after an scam or a ring doorbell, very very
similar idea. From a user's perspective, it's almost the same
(08:47):
as a nestcam or ring doorbell that says, somebody's stealing
your package.
Speaker 2 (08:50):
Here's a video.
Speaker 3 (08:51):
And actually, I think it's a big part of why
the product's been so popular and being widely adopted because
the customers actually are used to this concept. To the
customer seems really simple. Behind the scenes, the whole system
is incredibly complicated. So it starts with a piece of
equipment that we design a manufacturer called a Pane station.
This includes about forty components, including two high definition security cameras.
(09:15):
These are off the shelf top of the line six
megapixel security cameras designed for ruggedized environments, and about forty
other components. We assemble them in our factory in San Francisco.
The systems also include an edge computer which has logic
on how we control the cameras. We also include networking
equipment so we can send the data up to the
cloud over cellular broadband connectivity. We have power management. Sometimes
(09:40):
we need to include a backup battery. Sometimes we need
to include solar panels. Every site is a little bit different,
so it's a configured to order supply chain, which is
not for the faint of heart.
Speaker 2 (09:50):
And we mount.
Speaker 3 (09:51):
These systems typically on existing structures like cell towers, water tanks,
government communications towers, sometimes even private home homes or chairlifts
at a ski resort. We get really creative on where
to put these and we've actually found five G to
be a really great technology for us. One of our
investors is t Mobile, and they've been a phenomenal partner
(10:12):
in terms of allowing us to migrate from four G
to five G, and we're one of the first industrial
users of five G technology, which has further range into
the forest and higher performance than four G. So Once
we mount these systems, we rotate the cameras three hundred
and sixty degrees every minute, and we take ten frames
with each rotation, and we send this up to the cloud.
(10:35):
We put these images through a computer vision deep learning
algorithm and it draws bounding boxes on the images where
it thinks there's smoke. Each one of these bounding boxes
is reviewed by a human analyst in our Paneo Intelligence
center to make sure that it's really truly smoke. This
is called human in the loop AI. It's actually a
(10:55):
very common approach with computer vision AI because it's so
easy for a human to use damage as ground truth
to be sure.
Speaker 2 (11:02):
In that way, when.
Speaker 3 (11:03):
We trigger an alert out to our customer, we have
almost perfect signal to noise ratio, so the customers know
when they get an alert from PANO, this is going
to be smoke, and I don't have to worry about
being spammed. Then the customers get all the other rich
information they need to go control the fire.
Speaker 1 (11:18):
I'm glad you brought up the human loop element because
anyone who's followed any sort of AI, even things like
image recognition, you realize that there are outlier cases where
I think of it as like Paradelia where you're looking
up at the sky and you see a cloud and
it is in the shape of something, and as a person,
you know that's not actually the thing. It's not really
(11:40):
a cat up in the sky. It just looks like one,
but to an image recognition algorithm, it might fool it
into thinking it's a cat. And so having that human
element to be that step before sending messages out, before
you start to see agencies commit resources toward going to
a potentially remote location that is not easy to get to,
(12:00):
which could be taking their attention away from other very real,
very present emergencies.
Speaker 2 (12:06):
Yeah.
Speaker 3 (12:06):
Absolutely, And the good news is cameras as an AI
sensor are really phenomenal. And that's actually a core part
of Pano's long term vision and strategy is we've decided
to specialize in computer vision AI. We just hired an
amazing VP of engineering. He has a PhD in computer vision,
and we brought on a leader with computer vision and
(12:28):
expertise to lead our engineering effort because we think that
camera sensors are going to be the most important for
emergency management, whether those cameras are stationary like our first product,
or whether those cameras are on a drone or satellites.
With camera data, you can run your algorithm and you
do object detection to detect the signal you're trying to detect,
(12:50):
but you can also have a human verify in real time.
Because the camera data is ground truth, you can combine
artificial intelligence and human intelligence together in real time with
camera data beautifully, and so that makes it, I think,
much more useful and much more powerful.
Speaker 1 (13:12):
You also mentioned five G connectivity, which is obviously something
we really like to talk about on this show, and
it's interesting because you were bringing up a different sort
of flavor of five G than what I often talk about.
I'm often talking about the ultra high frequency five G,
where you have very high throughput, very low latency. But
for those who do not know, there are other bands
(13:34):
of five G, and as you were eluding, one of
those bands is one where you are able to get
much further broadcast range than you would with other cellular technologies.
So you mentioned about the partnership with T Mobile. Can
you talk a little bit more about that and how
five G has played a part in pano AI's operation.
Speaker 3 (13:54):
Yes, absolutely, so, we've been partnering with T Mobile for
over a year now. I think we were in the
twenty twenty two class of the five G Open Innovation
Lab and they help facilitate a partnership with T Mobile
over a year ago, and that has been a really
successful partnership. T Mobile helped us with technical support that
allowed us to migrate from four G to five G
(14:17):
sooner than almost any other industrial application, and this has
really been terrific for us, and I think it's just
going to continue to be more powerful over time as
we unlock more and more features that take advantage of
the performance of five G. So the benefits for our
system of five G are, first of all, the range
as you referred to. We're installing our PANA stations on
(14:40):
mountaintop locations, typically at the wildland urban interface or deep
in the forest. That's where you need to have wildfire cameras.
You don't need to have wildfare cameras right and downtown.
And sometimes our locations are not very close to a
cell tower. They can be far from a cell tower.
We could be mounting on an abandoned four service lookout
tower that's miles into the forest. So the fact that
(15:00):
five G provides for further range from the cell tower
means that we might be able to use cellular in
a location where otherwise cellular would not be an option.
Second of all, the latency and the broadcast and the
upload speeds and the bandwidth are actually very beneficial for
us as well. I mean, remember we're an upload application,
not a download application in terms of latency. This really
(15:21):
comes into play when we're using the cameras in optical
zoom mode. So our standard operating mode is patrol mode.
We're rotating the cameras one rotation per minute, scanning the
landscape looking for smoke. But once we find smoke, then
either Pano or our customers will take one of those
two cameras out of rotation and zoom in on the
(15:44):
smoke using the very powerful optical zoom capabilities of the camera.
And this actually provides much richer visual information of what's
going on with the fire. It allows the fire agencies
to much more clearly see road access, to see which
way the wind is blowing, to see strokes nearby. It
can really have a huge impact on firefighter safety. And
(16:05):
the challenge is as you're going to navigate the camera,
if you're on a four G connection, you have to
wait to send the instructions out to the camera, and
it can be a nuisance and slow down the navigation
and take you longer to find the fire and zoom
in to where you want to get to. It's not
the same as performing open heart surgery, but the latency
really will be of great advantage. And then finally, in
terms of the upload bandwidth, these cameras are capable of
(16:28):
thirty frame per second video, but on four G we're
typically uploading about ten frames a minute. Depending on the
four G connection we can go a little bit above that,
but with five G we can go to the full
thirty frame per second video, which, again, when you're staring
at a fire, you want all the visual information you
can get.
Speaker 1 (16:45):
Yeah, that's incredible. Well, can you tell me a little
bit more about the technologies that are on the back end.
I'm assuming that you have a mixture of cloud computing
and on premises compute that are processing all this images.
Can you kind of give us a big picture of
how that works.
Speaker 2 (17:00):
Sure.
Speaker 3 (17:00):
I love the theme that a technology is developed by
one industry and then other sectors find ways to harvest
that technology and use it in new ways. So at
PANO we're harnessing many technologies that were developed for other industries.
Five G is a perfect example, was developed primarily for
cell phone users, but we're using it for wildfire detection.
(17:22):
The AI that we use was really primarily developed for
the self driving car industry, and we're harnessing that for
fire detection. Satellite technology is really critical for us. We're
using geostationary satellite data that was developed by Noah, the
National Weather Service. Those have been very useful as a
supplemental detection method using thermal sensors. We're harnessing cloud tools.
(17:46):
I mean we ingest massive amounts of data. If we
were not able to harness scalable cloud technologies and Kubernetes,
there's no way we could do what we're doing today.
We use SaaS tools like Twilio to push out our
notification and scalable fashion. So there's dozens of modern day
technologies that were developed over the past ten years that
(18:07):
are included in our solution that our customers are totally
unaware of don't have to think about as Our job
is to go shop for the best technologies like five
G and bring them to our customers as as soon
as they become readily available.
Speaker 1 (18:19):
I love that explanation as well. It is another inspiring
approach to this and the idea of using that specifically
for a climate oriented task here where we know for
a fact that we're going to have more of these
climate events. It's interesting to me about your business because
when I first heard about what PANO does, to me,
(18:41):
it was immediately one of those things where I thought, Oh,
that sounds like something that you would think of as
an adjunct to a government agency for example. Can you
talk a bit about what thinking went into making it
a business, how did that sort of blossom?
Speaker 3 (18:57):
So it absolutely is a service that should be provided
by government agencies, so you're absolutely right there. And then
government agencies typically select vendors by running an RFP, and
that's how they select the vendors that then deliver the
government services that you're used to. And actually we've been
very impressed with the rigor that government agencies have used
(19:17):
when selecting vendors.
Speaker 1 (19:20):
Excellent. Can you talk a bit about sort of the
skills that are needed for a business leader who wants
to work closely with governments? What sort of skill set
does that person need and what sort of expectations should
they have?
Speaker 3 (19:33):
Yeah, so I had worked in government, but selling into
government was not my skill set. The first person who
joined our team, I'm very fortunate, is our chief commercial officer,
Arvinsattiam and he did have ten years of experience selling
into government. He had been working at Cisco for many years,
and for the past ten of those years he was
one of the leaders of their government IoT business, where
(19:56):
he sold million dollar deals into sy like the city
of Barcelona that led to wiring up the entire city
to be a smart city with smart street lamps and
smart parking meter, smart trash cans, and so he is
exceptional at learning how to navigate the challenges of selling
into government buyers. So I would say that selling into
(20:17):
government we have found to be more difficult than selling
into any other segment. So you need both sellers and
you also need regulatory We have regulatory advisors as well.
But I will also say a really important part of
Pano's business success has been that we don't only sell
to government. So our business model is that we deploy
(20:37):
in a region and we sell subscriptions to multiple types
of entities. So we sell to government agencies city, county, state, federal,
but we also sell to power utilities. They're one of
our largest customer segments. They have a lot of wildfire risk.
Their power lines run through high fire risk areas and
so they're really focused on protecting their ass that's we
(21:00):
also sell to other property owners like ski resorts, the
forestry industry, other private landowners, real estate developers. So it's
been really important as a venture backed company that we
have a mixture of public sector and private sector customers
in order to meet the business objectives. And also actually
(21:21):
we found these two types of buyers to be complementary.
We found that our private sector customers appreciate the fact
that we also have a public sector business. In fact,
we've done multiple deals where we've actually signed up both
a public sector and a private sector customer at the
same time. For example, in the Telluride area, we signed
up the Telleride Fire Department and SA Miguel Power Company
(21:44):
as well at the same time for the pilot. This
is called a public private partnership and both parties feel
great about this collaboration, and it's because the fire agency
feels great that the utility is participating and contributing to
support protecting their assets, and the power utilities know that
if the fire agencies are participating as customers that they're
(22:07):
going to use this product because the enterprise customers can't
suppress the fire on their own. The end users of
the product are the fire agencies.
Speaker 1 (22:14):
It's really interesting. It also makes me think of cases
where there are issues with power utilities and their close
association with fires. I'm reminded of rolling brownouts that were
needed in times of high winds, for example, where there
was a concern that utility lines could potentially cause fires,
and potentially down the line, a company like PANO could
(22:37):
end up delivering lots of information that could lead to
things like infrastructure overhaul. Like it's interesting to me how
your role could expand beyond the immediate detection of fires
but also ultimately contribute to improving infrastructure in places that
need it. And it's because you're able to gather the
information that's necessary before anyone can make those kinds of decisions.
Speaker 3 (23:01):
Yes, you're spot on. I mean, once again, I think
you know our long term vision better than we do exactly.
You know, our first solution at PANO is focused on
delivering real time, actionable intelligence during the response phase of
an incident. And by the way, I should also mention
our customers use our solution during fire season for enabling
(23:22):
rapid suppression of fires, but they also use our solution
in the rainy season for enabling safer and more manageable
controlled burning. In fire season, our solution is used for
pushing out real time actionable information to help with suppression,
and that's called the response phase of a natural disaster.
But there's three other phases of natural disaster management. There's
(23:44):
the recovery phase. There's the mitigation phase, which is hardening
your system, which would be for example, decisions around bearing
power lines or relocating assets, for example, making a decision
that you're going to change your deck material to make
it non flammable.
Speaker 2 (23:59):
That would all be part of mitigation.
Speaker 3 (24:01):
And then there's preparedness, which is okay, making sure you
have your evacuation routes planned out, making sure you have
a way to contact people, making sure you have shelters
with enough supplies. That's all preparedness, and then all of
that feeds back into the response and the cycle continues.
To your point, the data that we are creating we
create in our tool. After we push out an incident,
(24:22):
that data doesn't just go away. We've created an archive
of all of these fire incidents that we've alerted on,
so we have a very accurate record of exactly where
and when the fire started, and we ingest third party
data as well that allows us to have data on
how that fire was responded to. We ingest all of
that into our database, and we now have a treasure
(24:45):
trove of fire activity fire response all over the country.
And as our deployments increase, that data is just growing
and growing. And when I think about the mission of
PANO and the impact that we can have, I see
as twofold. One impact we can have is the impact
of improving real time response and lessening the harm of
wildfires in the immediate term, but I believe that we
(25:06):
can have additional impact by improving hardening decisions. There's going
to be trillions of dollars of investment made to harden
our infrastructure and rebuild our cities, rebuild our transportation in
the face of climate change. It's going to cost billions
of dollars to bury power lines to create microgrids. That's
(25:27):
going to take decades, and we're going to need to
spend billions for new helicopters and new planes. How are
we going to decide how to prioritize those investments, and
which investments are needed and which ones are not needed.
All of that needs data. There's no data today. There's
a federal agency that created a national Fire Incident Database,
and they created a database with ninety six fields that
(25:49):
should be populated for every fire incident. And when you
look at the database, only two of the fields are populated.
One field is when was the fire called in and
how many acres was it when it was And even
those two fields are suspect, right, how do we know
if that's really accurate information? But with Panos technology, you
can absolutely get that information accurately and many, many more
(26:11):
of those ninety six fields can be populated as well,
and data can really be an enabler for us to
adapt to climate change.
Speaker 1 (26:18):
Sonya, You've talked about so much here with the Internet
of things, artificial intelligence, the era of big data, the
ability for us to be able to do something meaningful
with that data. I think a lot of these things
all started to converge in a really exciting way around
the same time. The future you're describing as one that
I expect we're going to see lots of other businesses
(26:38):
focus on. I hope to see more organizations take that
approach of how can we leverage the massive amounts of
information we collect in ways that have a real positive
outcome in the real world.
Speaker 2 (26:54):
Absolutely, you're spot on.
Speaker 3 (26:55):
I think there's a hopeful message here, which is that
we as humanity have been work working on trying to
find solutions to slow down or reverse climate change for decades,
and I think many of those I'm optimistic about, and
I think human innovation will find a way to reverse
climate change, but it's going to take decades. There has
been almost no innovation that has gone into adaptation to
(27:17):
climate change, and as a result, what that means is
that there's so much low hanging fruit. We have so
many technology tools that are perfectly suited to help cope
with the effects of climate change, sensors, drones, satellites, five G,
big data software tools, and the entrepreneurship wave of the
(27:38):
climate adaptation field is just beginning. Panel is one of
the first companies in this field, and we're hoping to
blaze a trail here and we're hoping other entrepreneurs will
see that technology really can make climate change less harmful
and that we can save lives we can save homes,
we can save trees, we can save animal life. Let's
put our heads together and think about ways that we
(27:59):
can innovate to lessen the harms of a climate change.
Speaker 2 (28:02):
It's already here today.
Speaker 1 (28:05):
I had a couple more questions to ask Sonia before
I could let her go. Well, next up, what is
the best piece of advice you have ever received?
Speaker 2 (28:19):
A professor of mine?
Speaker 3 (28:20):
In college senior year, I was taking a behavioral economics class,
and he was explaining that people do not do a
good job of intuitively understanding probabilities. So people often are
afraid of losing something, even if it's a small amount
or a low risk of losing. People are grasped what
they have. They don't want to lose something. They buy
(28:41):
lottery tickets because they think they have a chance of
winning something big. So what he encouraged us is take small,
positive expected value risks in life. And I think that
was really good advice.
Speaker 1 (28:53):
I think so too. Yeah, I like that a lot.
Generally speaking, humans are really bad at assessing risk unless
we put a lot of effort toward it, I find. Yeah,
what does the term restless ones mean to you?
Speaker 3 (29:07):
You know? I love the name of this podcast, The
Restless Ones, and it made me think of not being
satisfied with accepting the world for how it is, but
saying I don't like how the world is today. I
think the world should be different, and I'm going to
go do something about it.
Speaker 1 (29:23):
Excellent. I couldn't agree more. Thanks again to Sonia Castner
of pano Ai for talking with us on The Restless Ones.
Sonia's story reminded me that innovation isn't restricted to inventing
the next light bulb. Innovation can involve finding a new
(29:43):
way to use that light bulb to tackle tough problems.
Taking advantage of advancements in technologies ranging from IoT deployments,
image capture and artificial intelligence, big data analysis, and yes,
advanced wireless technologies makes pano Ai a true pioneer. Moreover,
I think Sonya's story reminds us that in the face
of huge challenges, there is opportunity, and if we can
(30:07):
look at climate change as an opportunity to dedicate engineering
and innovation to protecting ourselves and future generations and with
any luck, eventually reverse some of those changes, that's truly inspiring. Again,
thank you Sonia Kastner, and thank you listeners for checking
(30:29):
out the Restless Ones, be sure to look through our
back catalog of conversations with leaders at the intersection of
business and tech, and come back for more great discussions
with the future Restless Ones. I'm Jonathan Strickland.
Popular Podcasts
Good Game with Sarah Spain
Good Game is your one-stop shop for the biggest stories in women’s sports. Every day, host Sarah Spain gives you the stories, stakes, stars and stats to keep up with your favorite women’s teams, leagues and athletes. Through thoughtful insight, witty banter, and an all around good time, Sarah and friends break down the latest news, talk about the games you can’t miss, and debate the issues of the day. Don’t miss interviews with the people of the moment, whether they be athletes, coaches, reporters, or celebrity fans.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations.
Crime Junkie
If you can never get enough true crime... Congratulations, you’ve found your people.