Alex Foessel, Market Technology Strategy Specialist, Pittsburgh, PA

There seems to be no end to the use of AI Innovation. Alex Foessel shares his view. Managing Partner at Balanced Engineering, a Market Technology Strategy Specialist, Ag Innovation and Off-Road and Connercial Vehicle Automation Expert. https://www.balanced.llc https://www.linkedin.com/in/afoessel

Episode Transcript

(00:02):
Welcome to Talk to Brazil with TomRioch, the business connector to business in
Brazil. Talk to Brazil is aleading business podcast talks to business experts throughout
the world. I'm Tom Rioch,known as the King of Networking, connecting
people from my studio in Brazil.Joining us again today from Pittsburgh, Pennsylvania,
Alex Fosel. He's managing partner atBalanced Engineering. He's a market technology

(00:28):
strategy specialist, AG innovation and offthe road and commercial vehicle automation expert.
Alex, thanks for being here again. The last time we spoke, we
spoke about AI in off the roadequipment. There seems to be no end
in the use of AI and innovationright, not only in equipment, but

(00:49):
in everything. AI and AGRO isnot only in machines. So are we
still at the beginning of the curve? Good morning, Tom. First of
all, thank you for having meagain. Yes, we're just at the
beginning of the curve. And thatis because the amount of possibilities of using

(01:12):
artificial intelligence for not only the equipment, but the business of growers and also
for all of the other businesses thatare part of the agriculture ecosystem or I'm
going to say more broadly of roadapplications, including construction, mining, and
others, is endless. I justcame back this week from another conference from

(01:38):
the Association of Equipment Manufacturers that includescompanies that build machines from all over the
world. You will find their Comatsuand Cummings and the Caboda and John Deere
and Case, New Holland Aco andmany many others. Right, and this

(01:59):
Associate creation has this conference called ProductSafety and Stewardship ah h h. And
one of the key things is thatthis year, for the first time,
we had multiple talks focused on artificialintelligence. Last year, I believe that

(02:20):
our talk was just the only onethat had the title AI, right,
because we wanted to bring awareness thisyear in addition to our talk that is
not only talking about and bringing awareness, but we wanted to present some of
the progress in creating tools to betterintegrate artificial intelligent composers. When we talked

(02:44):
last year, it was exactly theinception of AI. But you did mention
we did cover the part of safety. But just remind our listeners, we're
talking about the off road equipment.You're talking also of equipment that's driverless it's
autonomous, and so we're talking aboutsafety. It's not the safety of the

(03:04):
driver, the pilet whoever, it'sthe safety in the system that's running a
machine that doesn't have anybody in it. Correct, Correct, And that's thank
you for the reminder the use ofartificial intelligence. And by the way,
artificial intelligence is a phrase that hasbeen so used in so many contexts,

(03:27):
that is, it means something differentto every person. I agree. So
maybe one of the first things thatI would like is that I heard one
of the first presentations during the conferenceabout AI, but was much more in
the area of how do you facethis explosion of services and platforms related to

(03:53):
generative generative AI. And I'm thinkingof chat, tiptv in the most known,
but many many other platforms that areproviding services to create visual images,
videos, etc. And by theway, I would like to point then
later to another use that we havefound fantastic, very very good for some

(04:15):
of these tools. That is avery practical application in training. But coming
back to the conference, Okay,so Bruce Rasa presents these tools in how
you can compute compose music, andyou can write text, and you can
do analytics, and there's all ofthese applications from a business perspective. Oh,

(04:39):
he said, Bruce, I'd liketo ask you a favor. Would
you authorize me to pick one ofyour slides, one that defines terminology,
and I'm going to attach it tomy presentation so we can actually provide some
clarity to the audience. Well,you were talking about this AI, we

(05:00):
were going to be talking about thisother AI, which is much more product
related, right, Because I believethat it is our responsibility, Tom,
for those of us working in thearea to provide clarity, to reduce ambiguity,
and to explain to the audiences orwhenever we can have a conversation that
AI is a very very broad fieldand we need to maybe focus a little

(05:26):
bit more in the different manifestations ofit so we can address that more effect
I agree. I can imagine.And you mentioned all of the manufacturers that
were present, all right, thatare today manufacturing equipment. So my question
to you, obviously, when wehave conferences like this, when we have
associations that get everybody around the tableat one point, you're trying to reach

(05:50):
for commonality. Obviously, everybody hastheir own strategy, commercial strategy. They
want their own innovation. But whenwe're talking about safety, the safety needs
to be inherent of everything. Sothere should be a need for some type
of standardization amongst all of the differentproducts. And it's correct, and that
is one of the topics that wewere discussing, because, for example,

(06:14):
I could build a machine that isgoing to be significantly more effective. I
can build a machine that will useless fuel, one that is going to
be maybe more intuitive to interact with, and I can sell to the market
a this is how my product isbetter, faster, competitive edge. But
I'm never it's very difficult that I'mgoing to come out and expose myself and

(06:38):
say my product, by far isthe safest in the industry, because if
something were to go wrong, youwould be in a pretty difficult place.
That's why manufacturers get together and they, as you will said, create standards.
The challenge is that the standards thatwere created to produce to engineer developed

(07:00):
safe machines TOM are becoming to someextent a bit obsolete. Good news is
that there is a lot of workhappening and actually my business partner requires is
part of these conversations to update standardsand also to produce new standards. That
are going to be tackling. Okay, how are we going to build safe

(07:20):
robotic equipment? So then farmers,growers, bystanders, mechanics, the person
that drives the truck with fuel,whoever is working around machines, whether in
construction work sites or in farm fields, they will know that they can approach

(07:42):
a robotic tractor, combine, etc. Or an excavator and they will be
able to very easily understand in whichstate that machine is and also what the
machine is going to go do nextto The more we standardize that language or
signage or lights or sounds between machinesand humans, we're going to have significantly

(08:09):
safer and effective interaction and collaboration betweenprobotic machines and people. Well, your
explanation is clear. My question isalong the line, have all these manufacturers
bought into that? Does everybody believethat's the way, And that's a good
question. I believe. I believethat you ask them, they will say
yes, of course we buy intothat. The question is do we all

(08:31):
understand the same things? That iswhat I was referring to. That first,
we need to be very clear andvery precise in how we use language,
so we can understand what is itthat we're talking about, because if
you think of let's say, onething is to make sure that a module
that has a neural network that iswell trained to detect bystanders, so the

(08:56):
machine stop. It's very different from, for example, the analytics agents story
I'm using a technical word, butthe analytics software that is going to be
analyzing, for example, the agronomicswhether et cetera, to recommend what and

(09:16):
where to apply chemicals in a funBut it's also AI, and we can
say both are AI, but theimplications in terms of safety are very different.
Both of them very important, right, and then of course you have
many other things, like you know, there is now more and more AI
integrated into financial tools for running it. And what you're saying is absolutely amazing

(09:43):
because this coming week now we havethe AGRA show here in Brazili and Ribereto
which at the center, yes,the exposition of machines and whatever. Just
today this morning, I just sawan article of the use of drones and
many let's say farm equipment manufacturers,and I'll turn to manufacturing drones. So

(10:05):
where you may have a equipment thatused to apply pesticides on wheels or tracks,
an now they're all flying around andso when you get that, one
thing is that it's on a trackor on wheels and it's going through the
field. But now you can flythem over everywhere and you can apply pesticides

(10:30):
or fertilizers or whatever, and sothat's a whole different challenge. I said,
Wow, you know, I'm stillfrom the nuts and bolts times where
you talk about standardization, you weretalking about alloys, you were talking about
hard things that you can see andtest. Today I'm not even sure how
you can test all these things.Oh, Tom, absolutely, So first

(10:54):
of all, just a side comment. Certainly drones are becoming used application,
but one thing that we need toconsider is that today the operational deployment of
drones to do application is limited comparedto the coverage that the larger machines do.
However, we have areas, andBrazil has much of that. Areas

(11:16):
that have are very slope fills orareas that require precise application, and drones
will are finding the right places tobe used. That is exactly the example
that I saw, really applying apesticide and fifteen square meters on the corner

(11:37):
of a field because you're a machinecoin get there or didn't or whatever,
and so you send your drowne outand a couple of squirts and you're done.
And also drones have the fantastic abilityto map real fast. And you
know, companies like Haysen and areusing drones to map where the king fields

(12:00):
have gaps so that then they candirect the equipment that is going to plant
cain in just those areas to havea very productive cane field. But back
to our aies and everything AI inthat. Yes, but actually let me
address a real quick one topic youjust brought, and that is how do

(12:22):
we test How do we know that, for example, a neural network has
been trained correctly so that we knowthat it's going to be effective at detecting
bystanders. That is exactly what someof us are, you know, just
focused to understand what are the engineeringsteps, what are the best practices to

(12:48):
define the type of network, togather the training data, to gather the
test data, and then to testthis in a very methodical way so we
understand, well, is it effectiveat detecting, for example, children in
a crop. Okay, we're goingto test that, and we use tools

(13:09):
such as we contaminate the images alittle bit. We make them more diffuse
so that we see is the AIstill detecting that that kid. If it
is, it's great, But ifit's not, then we know to need
to go back and retrain the networkor do something in order to strengthen that
ability to detect children. So themachines can protect people. That's one example.

(13:33):
So and as to the commitment,every company, every equipment company is
committed to safety. The challenge thatwe have now is that we have newer
technologies and we don't have the tools. But rest assured Tom that many of
us are just doubling down our focusto conversation. And I see that on

(13:56):
your site from Balanced Engineering that's yourfocus. I understand that it's a challenge.
You need energy, you need tounderstand, you need intelligence, you
need real intelligence to help understand whatthe artificial intelligence is involved in all of
that. So from a company wideview and from Balanced Engineering, one of

(14:20):
the things I see then that youtry to find advanced safety solutions. So
you're really keying in on that,right, correct, Yes, and this
is for now applied to equipment tomachines, right. But we have colleagues
working in automotive. We have colleagesworking in AI and by the way,

(14:41):
appreciate reading that note about the UNdeclaration stating that there is a need for
frameworks right even though all of aisare maybe different, but at the end
of the day, we're still talkingabout the safety and well being of us
humans. It well, Alex,I want to thank you for coming.
We're running the end of our time, but again I want to share with

(15:05):
our listeners. We both are membersof the Future Advisory Board here in Campionis.
You're in Pittsburgh, You've got along productive life here in Brazil.
We've met and I want to lookforward to meeting you again when you're back
here at the upcoming Campenis Innovation Weekin June, so I hope to see
you there and we can talk moreabout artificial intelligence. Oh Tom, I'm

(15:30):
really really excited to see Campenis continueto stand up and bring more and more
focus on all of the innovation thatis happening in Campenis. And you can
count that I will be there andthey'll be very happy to talk about the
multiple tracks and the multiple topics thatare going to be covered during that fantastic
event. You can be sure we'regoing to be talking before that. Okay,
So where can where can our listenersfind you? Well, first of

(15:56):
all with your podcasts, but thenof course only Alex Fossil and if not
always you can write to me atAlex dot Fossil f O E S S
E L at the Balanced dot ll C. Very good. Well,
thanks again for being here, Tom, I have a great day. Thank
you for talking to me and forour listeners. It's Alex A. L

(16:18):
e X the last name f OE S S E L. You find
him at Balance dot l l C. Doc to Brazil's brought to us by
FOCUSMI Market Intelligence, an agricultural researchspecialist here in the Brazilian agricultural market.
More about them on their site fO c U S m I dot com.

(16:38):
Remember when you talk to Tom,you talk to the world. Goodbye
and thanks for listening. Thanks forlistening to Tom Riok on Talk to Brazil,
The Business Connector to Brazil