Business as Usual: Bloomfield Robotics

 

Transcription: 

So good afternoon, everyone. This is Audrey Russo kicking off a Monday. And this is business as usual. I'm president and CEO, the Pittsburgh Technology Council, joined by Jonathan kersting is vice president of all things media and marketing at the tech Council. And thanks to Brian Kennedy for being our executive producer, and making sure that we have great guests, and today is no exception. But before we jump into today's guest, who I've known for quite a while, I just want to thank Huntington bank, I want to thank them for believing in us right from the onset of COVID. And working with us as we experiment and pivot and make sure that we're connecting the community the reason why we do today and every day is because our intent is to make sure that the tech ecosystem is connected, and that there's opportunities for some pathways because you know, the show must go on business must go on and connections must happen. In addition to that, we really have an opportunity to talk to most of the rock stars that are in this region, and today is no exception. And so if you don't know anything about Huntington bank, they did help us through the pandemic. They help all you know, many small businesses, they were really involved in making sure that the support is there. You don't know them, get to know them. We have a chat. The chat is going to allow you to ask some questions. Hopefully we'll have time with Mark. To be able to have some questions answered. This is a this is considered a casual exchange. So we want to make sure that we have an opportunity for engagement and then we've muted your microphone so that way, we don't have to hear everything in the background. However, just full disclaimer, if you hear dogs in the background on mine, you'll know that I didn't mute my microphone. We have a great week ahead. We also have tomorrow. We have the project lead of the circular economy. It's Vesterbro. We also have the CEO of rich rivers agile. On Wednesday, Thursday, we have Deb bachlin, president and CEO of QED multimedia, they want a bunch of Emmys, lots of things to talk about and Friday with Finch Fulton, Deputy Assistant Secretary of Transportation policy for the US Department of Transportation, so busy packed week. But I want to jump right in because I'm very excited to talk to someone who I think of as a friend, as a colleague and someone who's really worked hard to build the entrepreneurial ecosystem as well as his role in teaching students over at Carnegie Mellon. So with this is Mark de Santos. He's now the CEO of Bloomfield robotics. As I mentioned, I've known mark for many years, and in many capacities, he's easiest described, and Mark, you can tell me whether this is right, as a serial entrepreneur, but your interests and you know, your experience really are wide and deep. So from everything from early onset of doing radio, as well as your interest in public policy, I don't think there's anything that you haven't touched, you know, you're adjunct professor at Carnegie Mellon, in the College of Engineering, you are the Chair of the Board of allies for children. And before his life in academia, I mentioned that he was active in public policy. He actually worked for senator john Heinz, and as an analyst in the White House Office of Science and Technology and policy. So he's done a lot of things. And he has a lot of opinions. But really, what matters today is we're going to talk about his work in robotics. So Mark, welcome.

Thank you. Thanks for that. So I have to record that and send it to my mother. Mother, very proud of

me to hear that listen is recorded. So just so you know. Hopefully, she's proud of you. Yeah. And then yeah, then she can give me a little bit of a kickback. Okay. So really appreciate that. So Mark, we, you know, yeah, today, believe it or not, we've just kicked off our fall stem summit, which reaches about 400 plus teachers and students who have registered from around 250 schools and organizations in southwestern Pennsylvania. So, as I mentioned, you're currently serving as chair of allies for children.

So I'm still on I'm just on the board now. I'm just on the board.

Oh,

yeah. Talk about that organization.

Yeah, sure. So, yeah. So it's an organization is dedicated to supporting the development of children who, in the community that are not reached by public entities. You know, nonprofits, the point of a nonprofit right is to fill that gap where private and government don't fill the void. And so this organization was created. To fill that gap, and those gaps, and it's doing a great job, and it's a great team and Pittsburgh is, you know, it stands out as kind of a place where people take care of one another, one way or another and try to help one another. And this organization was really designed to kind of augment what are really public resources. And I think it's a good model for other sorts of organizations around the country, actually,

well then maybe share a little bit about your own early days and the influences in your own life, because I think that really has shaped a lot in terms of your career, like in entrepreneurship public. So

my earliest memory, actually, they've been kind of interested in maybe what I do now, which is tech type stuff, goes back to when I was a kid I actually grew up in was born here in Pittsburgh, but I grew up in, in Charlottesville, and where people know in Mercer County, hour and a half north of here. And I remember I was a I think it was maybe fifth grade, we went on a road trip to Pittsburgh, we went to the museum and the museum, there were these models of machines and all sorts of crazy things, catapults and bridges and all that made the design by Leonardo da Vinci. And they had built these models of all these different ideas that wouldn't models that Leonardo da Vinci had. And I remember being heavily influenced, I still have a vivid memory of seeing those, and thinking this guy lived 500 years ago. And many of these things were just centuries ahead of their times. And from that moment forward, I was always interested in, in in technology, I never became a technologist. So I've, but I've always been involved in it in some capacity, whether it's in policy sense in my way back when when I was in Washington, or now in starting companies. So I have a lot of respect for people that engineers and software engineers and developers and scientists, people that make things and invent things, I have a lot of respect for them. And I guess you could say my career has been around trying to create an environment to help bring that about, though, for me, that's what entrepreneurship is all about.

Well, you know, now at Bloomfield robotics, it's another great spin out from Carnegie Mellon, the technology, it's bringing artificial intelligence into the agricultural market. I've seen some great pictures of you out in fields and doing things that I probably never saw. Never saw pictures of you before. But you have these two products blinked in the flash, and you're already working in two key industries. Right.

Can Yeah, I could tell you,

yeah, talk about that. Yeah, it's

a it's sure. It's fascinating. Little known fact about Carnegie Mellon, among many things. Is that Carnegie Mellon's been in ag tech for a long time for decades. People think of robots they think of military applications, autonomous vehicles and variety of other things. CMU is, of course, well known for, but little known fact is the ag tech work that has been going on for years. And a guy named George Cantor. Dr. George Cantor is a research professor there came up with the idea of taking the perception piece, the ability to sort of look at stuff and carving that out as a service. So if you go back, you know, the agriculture is a, you know, 10,000 year, it's, you know, Mesopotamia, 10,000 years ago, they were growing crops. And from that time forward, proper inspection is largely a human activity. Somebody goes out among the fields and looks at the crops and then renders a judgment. Ideally, those people will trained to do that. Now you have drones, drones are now came on the scene and ag tech where you have drones flying over crops looking down. And they too are looking at the crops and they're looking for all the things that that individual is looking for. And typically, you know, these are agronomist, botanist, horticulturalists. viticulturalist looking at for things that tell you the condition of the crop. Something bad is happening when to harvest what the yield is going to be. That's a big issue predicting yield. But these guys with George and Tim Miller sim was a grad student of his but we're now founders said as we carve out the perception piece, let's use cameras. We'll take pictures of the plants. And then we'll use AI to look at those pictures to look for all the things that those horticulturalists bought agronomists look bored. And thus the idea was born. And that was about a year and a half ago, when the company was founded, I came on about a year ago. And they said, Mark, you know, when I met him, they said, We take pictures of plants. And we use those pictures to determine the health condition of the plant. And I was what took me about 15 seconds to say, Yeah, I want to do that. We are in. We're, we're crop agnostic. That's a phrase I never thought I'd ever use in my life. Never imagined I would ever say I'm cropping gnostic, but I am. But effectively, we don't care what kind of plant we look at. But we have focused and specialized right now because where the opportunities are, and the big one is in grapes. So we have our cameras, two types of cameras. One is a handheld device where you look up close, but the primary product we have, the service is rendered through cameras about yay big, I can show pictures if the audience is interested. And it has a stereo camera. So you actually use a two lens has its own light source, and then an image is five frames a second, and you just pass the camera in front of the plants. In our case, vineyards is our focus. And so this, these cameras often sit on tops of tractors, ATVs, anything that moves and you just pass it in front of the plant snaps away. And those pictures are then analyzed on the device and in the cloud. And we look for all the things that a

virologist viticulturalist would look for. And

grape size, density, grape, cluster, leaf, you name it. We are in a dozen vineyards. In four states. We're in several in New York, four in Washington State one in Oregon, several in California. These are paying customers, we are also about to go into two vineyards in France.

Two very large vineyards, you've heard up.

And we're looking at one in maybe Chile. So pretty exciting, pretty exciting. I never thought I'd be in the wine industry. But you know, we're deep into it, you know? So it's, it's crazy the world works.

Got some perks, it's getting some perks.

Not yet. Not yet. But I know I'll be you know, we can't fly to France, we have a we have a frog guy there in France, who we're working with who is our sort of person on the ground, who comes from that industry, we have a person in Washington State. We'll back to all that we work with who's our person in the West Coast. So we don't COVID is limit our ability to travel. But fortunately, we have people that we're working with to be the point to do the business end of what we do. It's exciting, just a statistic your audience should be aware of. So the population the world's about 7 billion expected to add another 3 billion in the next 30 years. In the next 30 years, the world will need to grow more food that has been grown in human history. And the next 30 years, and it'll have to be on the same arable land, in fact, probably less. And so you're basically saying, how do we get that prop load on the land? We have now and that's the challenge. I got a question. How should I answer the questions? As soon as I get

them, Jonathan can just prompt it. But if you read it, you can answer it.

Yeah, so a cert so can we see diseases virus? Yep. It's called integrated pest management. So when we look at a plant, we look at health, and we look at performance. Health relates to the things that determine whether or not the plant is going to survive. So we look for integrated pest management, disease, water stresses, all the things that visual observation would reveal we can also see so if the person can see it, we can see it. And if they can identify it, we can identify it as well. In fact, we have a partnership in that regard with Cornell University. Cornell has one of the premier viticulture programs in the world. They also have one of the largest, I guess a better way of saying is infested vineyards. So they intentionally infest a variety of grapes with all these different diseases and then they manifest visually, and then we our cameras can see it and then we can render a judgment about that plant. And some of these diseases are pretty exotic, so they don't often show up. The way AI works is you have to be able to see it to know, you have to be able to see it, train it and then see it again. So we have a great partnership there. So think of that as health. And then performance is, hey, this cluster should be this big the grape should be this dark. At this time of the year. Right now we're in something called Verizon. And so we have table grapes as well. Some of our customers are table grape customers, as well as buying grapes. We build our own cameras.

And,

you know, we don't these cameras, you can't buy them. So we're not really in the business of being into hardware experts. But the fact is, you can't buy these cameras to allow the AI work, sort of the time being we're making them but nope to be outsourcing that. In fact, somebody asked about Washington State, I'm actually speaking at a conference at Washington State a virtual conference here soon.

The fires have

had a severe impact on our customers, all the way up the coast, from what from Washington on down to California. So, you know, there's people I say, hey, when I had this, this, this cab and had a smoky flavor, well, this is a little bit of a little bit too much of a smoky flavor, if you will. So this is not the kind of smoky flavor people want. Also, there's, there's heavy metals and other things in that. So it may in many cases completely obliterate the grape, the crops for vineyards.

Wow. So what about on the marijuana side of the house?

Yeah. It's interesting, you asked about that particular interest you have in that

everyone who knows me knows I have a deep interest in that.

Somebody said, they said, Hey, Mark, you're in wine and cannabis. And I was gonna say, well, we, you know, slot machines be another one, we get that but um, so cannabis, we have cannabis customers, we have cannabis customers in Oregon, Canada and the Netherlands, actually, with a very large company, a very one of a gigantic company in the Netherlands, that makes equipment for indoor grows. The handheld device is the one we use. We're using cannabis, which basically, this is 10 times more than your audience wants to know. But we use something called a liquid lens. And so if you're taking a picture of a three dimensional object, typically you're getting one focal plane. So anything beyond that focal plane, or in front of it's a little out of focus. For a standard picture, like a cell phone, we can't see that. But if we're taking every pixel, like our AI is, it's out of focus. So this liquid lens, believe it or not, snaps multiple focal planes at the same time. And so this handheld device, you can put it up to a cannabis leaf, and you can see the trichomes, which are like the little hairs where all the good stuff is

the good side, like that technical definition.

That's awesome. You can see there's a couple of other questions here. Yeah,

I said, One, manage vertical. Yeah, mobile sensors, airborne hyperspectral. So my first company, which I made $6,000 on 12 years ago, was we would take hyperspectral cameras and put them into the belly of a Cessna. And then they would fly over a crop field. The challenge with using any kind of camera that's looking down on a crop is you can't see the the meaningful stuff, particularly for specialty crops. So just to differentiate, so if you took all the crops in the world, but 30% fall in the category specialty crops, simple way I would define that as anything you can buy and Whole Foods plus cannabis, some grapes and some other things, and flowers and so on. You have to be able to see those which means you have to get perpendicularly plant, you can't see apples, cherries, grapes from the air, you have to get boots on the ground, so to speak. So, um, so there'll be a place for hyperspectral cameras and other airborne sensing. Plenty of use there lots of big market, multi billion dollar market. But this emerging market of getting up and close personal to the plant is new. We have tests going on in two apple orchards in Washington State, two of the largest Apple growers in North America. We're about to do a deployment with one of the largest Apple growers in the UK. These are experiments, tests pilots, so we're getting into other crop verticals. And that's exciting. And for many of these crops you have to get on the ground. It kind of look at it.

Mm, do you have a couple of more questions? Right?

Yeah, yeah. So we've got

a, let's see is it next step sample crop me allies grapes is the machine with camera travels through vineyard. Yeah. So there's a, we were featured in good fruit growers magazine, again, another phrase I never thought I would ever use. But effectively, you basically put a camera on a ATV, I've even talked to an ATV manufacturer reached out to us that's trying to get in the ag market. And he's one of these guys on the ATV and go go 20 miles an hour, it's hard to do that in in a, you know, in a vineyard that you can go pretty fast and get a lot of data very fast. The entry point for our technology or the is the specialty crops, where the revenue per acre is much higher like a typical corn, you might generate 600 bucks an acre and in a high end, grape vineyard or tomato grow could be hundreds of thousands of dollars per acre. So you want to be where the where, you know, the expense is not substantial. And the demand is quite high.

And so what about clients in Pennsylvania

we don't have any yet. And it's not for lack of trying. It's just that we follow where you know, we follow we push on open doors and so we got inbounds from New York State three vineyards there for in Washington State and several acts in Oregon in California, we're adding more adding another one a big one on one of the oldest state venues in the United States as a customer, some of the customers are very well known. Tier One like in the two in France are considered the their world famous vineyard so the they have a very hot a

very high interest hops yet we've done hops.

We are working with

some folks to see if we what the opportunity is there. And I got an inbound from jack, crop yield management.

So one of the challenges

is maximizing yield. But there's another issue with quality. So you want to just not have as much yield as possible you want it has to meet a certain quality standard. And little known fact, but I you know how much I've learned about vineyards. I was thought you know, they go and they pick the grapes good to go. You know, the grapes are ready, go pick the grapes. And it doesn't work like that. They actually pick the grapes at different times. So they take grapes when they're ready. Um, and for some of these vineyards, it is high science, it is a very sophisticated operation as to which grapes are harvested at what time, even on the same line. And so the beautiful thing about our technology is when we look at a plant, we're actually geolocating that plant. So I have a motto sounds corny. Again, one, I'll say this one more time. And then that's it. But never thought I'd say this. Every plant matters, right? every plant matters. But for us, every plant matters. So literally, we look at every plant, we geolocate it. And so we follow that plant through its life. So we know that that's vine five, row seven. And believe it or not, for these vineyards, these vines, these individual vines are tracked. Now there's an interesting phenomenon that's happening in the in the in the in the Ag and that is that chain of custody. so crazy as it sounds, there'll be a day, not unlike when you buy an appliance, there'll be a day when you can determine who touched that tomato, specifically where that tomato came from, perhaps even to the individual plant, believe it or not, and track it through its life. Believe it or not, that's gonna happen. So the level of sophistication for is growing. We are part of a phenomenon that's been around for 30 years called precision agriculture. And precision agriculture is the idea that I closely measure all the stuff that goes into into a farm into a grow and then I manipulate those depending on what I'm trying to get out of it. And then from the time the crop is harvested. Forward, I track It's it's health. But for many farmers, the what happens in the grow from seed to harvest is still a little bit murky. That's where crop inspection comes in. This takes crop inspection to a level of precision that's unprecedented. And so this is particularly keen for like indoor grows. So you see a lot of that now. We're looking, we're talking two to two very large indoor grows right now.

So let me let me ask you something. Yeah, you get like this incredible amount of data that you're extracting in your Columbia. And you know, your centers are going nonstop, what and the volume of data is probably just insane. How are you? How do you see five g? In so this is this record?

This is great. Yeah, this is a great question. Andre, this is a great question. So the genius, my my George and Tim are very smart, and an affiliate hogestyn. Very smart. And what they said is, look, farms don't have fake big, thick Wi Fi pipes. So we can't just upload an exabyte you know, exit. So. So what they said is, the cameras have to be close to pure IoT device as possible. So a lot of that data is processed on the machine. So our cameras are actually having a great deal of processing power. And so they thin out the data load. And so very thin stream of data goes up to the cloud. So we don't assume anything about the farmer having some 5g connection. Now, I think 5g is important to two technologies I'll make in my, you know, extensive one year career in ag tech. With my vast experience of one year in ag tech, I'll make two comments.

Today, there have been waves of Ag tech,

hope and disappointment starting 30 years ago,

and particularly robotization of farms. And there have been enthusiastic disappointment, enthusiasm, disappointment, I think this time it's going to stick. And the reason is, ai finally works in a way that no one could have imagined even five years ago, the perception piece in particular, and 5g, those two things together make possible the dream of a vast, vastly more automated farm with much higher yields and much higher quality. It's not going to eliminate the farmer by any means farmers are under a minute immense pressure. It's a tough, tough business with thin margins. And there's a culture of hard work and commitment that that you do just see it. Every farmers eyes, they love what they do. They're committed to what they do. But they need tools to allow them to make money. Most agriculture in this country is subsidized to the tune of about $25 billion in here. We have cheap food because we as taxpayers are willing to fund that. That's why we have such inexpensive food, we pay tax money to go in, to subsidize that. So they're very thin margins, farmers need help these tools change the formula for a lot of farmers move farmers from being unprofitable, ideally someday to profitable. So they're going to augment farmer but hardly eliminate farmer.

So you have a couple of more questions. We're coming up for 30 minutes, but there are some really great questions out here. Is there one you want to grab?

Yeah, yeah, I like the stem question. There's some organizations were working with to try to encourage farmers people to think of farming in a new way. So my previous companies, you know, robotics, great company using similar technology to assess road services, they had a similar challenge with respect to stem in that, that technology is is you know, will and is changing civil engineering, at least road engineering by taking AI and allowing people who are managers of road networks and road services to, to be more effective. It's the same with in this business, it's this AI is not replacing the farmer or the crop inspector. It's augmenting the ability of that farmer and crop inspector to do something never done before. What that means is anybody who's going to be an agronomist, or a farmer, or go into that profession has to have science, computer science, and tech and information technology skills. They have To have it is it there is probably not a profession around not a inspection profession around that doesn't require some computer science and information technology skills, mainly because AI is creeping into every discipline, every inspection regime you can think of.

So, just one last question out of curiosity, because I don't know the answer to this. How did they come up with the name Bloomfield?

Anybody want to speculate? The answer to that one?

Okay. Marina Bloomfield, we totally agree. We interviewed the two, the two founders A few years ago, and it was all about Weber.

Weber plan word, so

I onwards.

I can. Yeah, it's a mark. Yeah, no, it's fun.

Natalie got questions here, will you? if people are interested? Can they reach out to you?

Yeah, absolutely. By all means, yeah. I,

there's a lot of things here that people have brought up mark that I think are ways that we all need to be helpful as one, you don't have any presence and doing any of this work across Pennsylvania. And when you think about this being an agricultural state, you got to figure out a way to make that to make that a reality. The second thing is, is how, you know, people are interested in marketing, and you know how you're getting the word out. So I guess there's a sense of pride in knowing that you're here and the work that you're doing. And then the last thing is, Are you hiring? Are there any hiring opportunities?

Yeah, we're interviewing. We're in the throes of raising more capital. But we are interviewing people anticipation of additional funding. we're generating revenue. We've been fortunate that we received, you know, over a million dollars in in grant funding, one of the more interesting projects, I'll just mention this for anybody out there interested in sorghum. So we have a project with the US Department of Energy. And specifically Clemson University within that program, and some program called Terra. And the US part of energy is interested in increasing crop yields for biomass and ethanol production. And so think of sorghum. And so we have developed the platform to the extent that one can. There's a field in Florence, South Carolina has 10 acre field, it has 100 varieties, over hundreds of varieties of sorghum. And if you look on our website, or on our LinkedIn, you'll see the page you'll see the cameras going through sorghum fields. And what we're looking for is the super sorghum. So each of the different types of varieties has a different genetic makeup developed by geneticists at Clemson. And then we're looking for what's called the phenotype or the genetic, physical expression of those genetic markers, and, and ideally, trying to find the perfect form of sorghum. And we do that with our imaging. So this is a product that would be of interest to breeders, you know, our focus is on growers. But there's an application where breeders can use this too. And but again, we've been fortunate to receive a great deal of grant money that allows us to, you know, get the company off the ground, we've raised some capital we're raising for now.

All right, well, this is Mark de Santos, it's a quick peek into the work that he's doing Linfield, robotics, you can hear that they're on a great trajectory. And, and actually solving some really hard problems, which we didn't get a chance to do a little bit of a dive in, when you started talking about the growth on this planet and the amount of food that is required and what that really means for the future. So maybe that's where we pick up the next time we talk to you about those issues, because I know all of us, you know, really care about that, and the where we get our food and the demand for food and what does it mean for local, you know, consumption and farm to table as well as you know, what does it mean for the world and I

may have, it seems, I may have some cases of some really, really good cognac too. But we got for free. So

that would be that would be even more amazing. But then we want to do a deep dive into the weed industry when

I guess we have to do all this in person as of just suggesting right after.

Yes, and toast to the days where we can do that in person. So everyone. That's a little peek into mark to Santas at Bloomfield robotics. I know he didn't get it. We didn't do a complete fleet service to understand who mark is the man but you understand the work he's done and the span that he has and that he is a serial entrepreneur as well as an academic and, and he's also a policy wonk. So Mark, the sanest, thank you so much for taking the time with us stay safe. And if anyone wants to reach out to him, you know where to

go. Martina it's Martin bloomfield.ai.

Market Bloomfield,

Mark. It's great to see you. Glad to say. Thanks again. Thank you all. Thanks.

Transcribed by https://otter.ai