 All right, good afternoon Thank you all for being here. My name is Nate Moser. I'm the department head and agricultural and biological engineering It's my great pleasure to introduce our next speaker this afternoon. I guess just barely afternoon dr. John jinn dr. John jinn was Comes comes to us and into a professorial role and kind of a unique Pathway and maybe he'll speak to that a little bit, but he's someone we recruited from industry who was very interested very practical problems in solving a very profound Global challenge right we have limited land we have changing weather patterns and we have growing human population So how do we continue to increase? agricultural productivity to continue feeding the human population and maintain our environment right do this sustainably into the future and so that was the heart of Former company that John was working for Pioneer so the working on trait development for the crops that we grow that provide the food that we ate for lunch today and We encouraged him to maybe have an opportunity here at Purdue University to think deeper about similar problems and how to apply His background his knowledge across multiple domains. I'm very happy. He has some mentors joining us here from the College of Agriculture He's worked very closely with on on solving real problems that are the main specific in the sciences that engineers such as professor jinn can Work on and and move us forward in solving these global problems. So please join me in welcoming dr. John jinn Well, thank you very much Nate. It's really my pleasure here to be here today It's a really exciting to see some of the collaborators college leaders and the young graduate students And I really want to say thank you especially to Marsha and Maria for coordinating this event So I'm a sensor developer for plant phenotyping And I realized that all the new associate the professors are thinking about the same thing when making the slides We will start from the map here. So So I'm showing the showing you the world map here Basically telling you the story of how I come from Asia to Europe and then to the United States And also how I came from industry to academia So I was originally born in Nantong China, which is a city across the Yangtze River from Shanghai Very close to Shanghai. So I use Shanghai airport to fly to Chicago. It's kind of the symmetric trip here and When I was 19 years old, I traveled not too far away 100 miles south of Shanghai to Hangzhou city and I studied the computer science in Zhejiang University So just for your information, Hangzhou is famous for this West Lake Many people say this is the most beautiful lake in China So I would recommend considering that because sometimes, you know, when you had a bad day sitting by the lake for about five minutes Everything is healed. So it's very very beautiful After my graduation, I traveled to Europe Seriously, the Little Mermaid was one of the major reasons I chose to travel to Denmark But of course we have a top ranked Technical University there. So I studied a computer engineering. So still in computer science or computer engineering so and after two years I Came to Iowa State University and studied for my PhD degree for agriculture engineering So I know this sounds a little unusual a student transferring from computer science, which is Top rated in paying but now coming to agriculture, but I would like to say this is one of the Wizzest decision I made in my life Basically, this gave me the opportunity to apply what I learned from computer science and the computer engineering in a huge market Which is agriculture? So after graduation after getting my degree I joined Dupont Pioneer the company's new name is Coteva So over there I was an imaging scientist So of course we worked on a lot of the projects, but the biggest project I worked on was the I led the development of that and the design of the imaging systems for Dupont pioneers 40 million dollars high throughput automatic phenotyping greenhouse facility So it was really a great opportunity for myself to gain some industry experience before coming to Purdue in 2015 So when I just came to Purdue I decided to continue the same work Which is to design the phenotyping systems and it was really exciting But also busy and hard-working first three years. So every year we delivered one phenotyping system so As you can see from the pictures the first one is that is my first project using conveyor systems to deliver the plants through the hyperspectral imaging station Automatically being scanned and we can predict nitrogen Or the nutrients disease and the water automatically and many of the hardware and the software IPs were leveraged into the College of Agriculture's AAPF facility, which is a second one The third picture is our second greenhouse project. So basically we built this Gantry the yellow gantry system to play the role of a mock drone system So we can simulate field remote sensing drone remote sensing in the greenhouse environment And then we actually transferred this system to the field, which is a last image So this is our field phenotyping gantry in our acre research farm So we use this facility to learn and we get got a lot of the knowledges about how for example, how the Diurnal changes and also the environmental condition variances Severely impact our phenotyping results and how we can actually build models to calibrate those So anyway, these are the big and the expensive facilities You know, we're very lucky at Purdue to have these facilities The one question is we also have so many plant scientists and we have 600 million farmers around the world How do we allow more people to benefit from the advanced phenotyping? Technology, so the idea is to put all the hardware and a software technologies into a small handheld device So they can be easy to use at any location for any type of applications So we made the first prototype product of Leafsback, which is the first Handheld hyperspectral leaf scanner. So as you can see from the video It takes about three seconds to scan the leaf and then from the smartphone You can read the prediction of nutrients disease water and each measurement is geo referenced and the leaf is a hyperspectral image So today, of course, it's not a research seminar. I just want to give you the big overview of this product So and and you can imagine that for this Leafsback. It gives a very informative Image about how the crops are growing Each pixel has hundreds of different colors and we can have one millimeter spatial resolution but this beautiful image doesn't make any sense to the farmers or to the plant scientists because we need to Predict the nutrients and the disease we need to get the data So that is another big part of my lab's work Which is to build all kinds of the image processing software and the models to allow people to just take one scan And immediately you can get the information about the nutrients drought, pathogen diseases, chemical impacts, and so on So the software part is another big part of my research Another third part is to develop robot technologies Basically, you know We are sitting in a very comfortable room here But if you go to the farm in the hot summer The temperature the moisture the sharp blade of the leaves It's really not friendly for human to stay in the farm So our hope is to develop robots to automatically Do the field scanning with our advanced sensors to replace the human being so for example, this is our new Drone-based robot called dr. Fino so it can fly to designated locations we program in the software and Once it landed at one location it it can automatically use the machine vision to detect the top matured leaf of The crop either on soybean or corn and then it use the robotic arm to drive the leaf spec Sensor to scan the leaf and again similarly you can see the results immediately from the smartphone app And it can fly to the next location and then the next location is actually optimized by the software itself Because every time we get the new data the whole nitrogen map of the field is updated And the software can actually decide on the best next location to optimize the efficiency of This field of scouting so these are kind of the small summary brief summary of my research So just to also want to share with you that industry relationship is very important for my research Actually most of my research funding is coming from industry just a couple of examples So for example, we developed the herbicide mode of action identification. So basically you give me one plant We can take a one scan and among dozens of different possibilities of different herbicide products We can actually tell which herbicide you sprayed based on the outlook of the leaves So this model is already adopted by Sumitomo chemical. It's a company in Japan working on chemical products And another second example is we also developed the wheat disease detection the early detection algorithm Which helps the company of FMC to double the daily test throughput because we can see the disease symptom Much much earlier than they are visible to human eyes so Commercialization is another part of my life now. So I was strongly Encouraged by by the university to commercialize some of the technologies. So We founded this company leaves back LLC in 2018 and this company won the price of Davidson price in 2021 So so this is just another big part of my life and as you know as a full-time faculty It's a big challenge for you to find enough time to really push the business going on Why are you still need to work on publication and teaching? So that is why I'm talking with a lot of the potential partners to hopefully push this forward We will see how this goes. So that's commercialization part Teaching so this is probably the one of the most different part of my career life here now Here compared with my industry industry job. So a lot of the memories if I Go back to four or five five or six years ago. So the first picture is actually my first class I created this new graduate level class AB 530 named the plant phenyl typing technologies so we're really enjoying a lot of the discussion and And at the end of the second semester, I taught this undergraduate class basically It's a circuit design and a robotic class for all of our biological engineering students so really a lot of fun to interact with the young people there and Actually, many of the students in the undergraduate class decided to follow Follow up with our lab and it became our graduate students later And the picture on the upper right corner. This is my lab Three years ago. So some students already left Purdue some of them join us So it's another big part of my life here at Purdue every day Interact with the students and we learn from each other and keep our projects going on. So This is a teaching part. So thinking about the next 10 years or 20 years First of all, I would like to share my dream So my biggest dream for my research is to close the loop of digital agriculture. So what does that mean? you know Agriculture is actually the least digitalized among all the major industries according to the McKenzie Global Institute digitalization index and also according to the McKenzie's report another separate report They expect the market for digital agriculture to be $50 billion by 2030, which is just seven years seven years later That is why investors are looking at the digital agriculture and try to invest and many many startup companies Were founded over the last years and so many drones are flying above the fields Actually This is not the slide actually I would like to go to this slide many many of the drones are flying around the field But we see one issue the issue is the farmers They are not willing to pay the money out of their own pockets to these technologies Taking Monsanto or Bayer's digital agriculture to the field of view many of you might know field of view is a Smartphone app field of view so far. Mostly. It's just like a happy box gift to farmers It's very rare for me to see that the farmers are really using the field of view data and information to make the decisions on When to fertilize how much to fertilize and when to spray the herbicide when to spray the Chemical products so in my personal opinion the bottleneck issue here is the sensor quality It's exactly because our phenotyping right now our phenotyping sensors We're suffering from a lot of the noises including the daylight changing the wind speed of the diurnal Change of the crops so among all the noise the signal is Embedded in is is buried underneath the noise So the quality is not good enough for the farmers to adopt these technologies and to make those decisions So but on the other hand I was also greatly encouraged by our recent data and experimental results So I think I'm seeing the light at the end of the tunnel That's the quality of the sensor is really improving and it's getting quite close to that level when we can close this loop so the farmers will really pick the data from our sensors and and Make the decisions for the field operations. So that is my biggest dream So in order to pursue this dream, of course, we need to keep the research going on But at the same time, I think the team is really important. So that is why You know over the last several years We had the plant science initiative 1.0 and 2.0 a lot of the very good collaboration Relationship has been formed and I think I'm seeing a great team here at Purdue So we have people from the engineering college having very good Technologies and also people from the data scientists team having very good technologies And also around our AB department south of the state Street We have so many the world's greatest plant scientists some of them are sitting here in this room So they know the application. They know farmers. They know what is needed. What is the problem in the field? so I think AB should play the role our Engineers that AB should play the role to connect the groups together and also we also would like to outreach to the outside of people including, you know the Industry people from the conversion center Purdue foundry and the Purdue OTC so to form a team to keep developing the new technologies and Push the tech technologies outside to impact the world finally So I'm thinking about maybe for the next 10 years 20 years I will be very happy to maybe spend some of the time to maybe do a little bit of the leadership work maybe through a center or through some additional collaboration projects to Work together with our colleagues to move this forward. So I think this is my Kind of the review and this is a really a good time for me to stop Think back and recharge and find the new directions for the future and I really appreciate your Attent of all of you attending this this meeting here. I will be very happy to answer any questions. Thank you Very good some questions Very nice demo Thank you really impressed you work The question is when you are trying to find out how much nitrogen you have in the plant What is the scientific mechanism that nitrogen can be tracked and not other elements? So this is a great question. There's a little technical so A brief answer is you know, generally the color of the leaf is a big reflector of nitrogen So nitrogen amount the amount of nitrogen actually directly Determine how much chlorophyll the plant can produce and the more chlorophyll it can produce You know chlorophyll is a reason for the leaf to be green So that is why we can take a picture of the leaf and by watching how green the leaf looks like We can determine how much nitrogen it is This is a basic level and this is actually the standard method. Everybody's applying However, I really like the second part of your question How to differentiate between nitrogen and the other nutrients like potassium phosphorus and others Because all the other nutrients including water even they all help the plants to keep green Lacking of all the other nutrients also Hurts the green color of the leaf and they becomes yellow. So that is why sometimes our existing models We claim this model is a predictor of nitrogen, but actually it is impacted by the other by the other nutrients deficiencies so a very But the hope I believe Stays in the still in the sensor Technology development so for example if we can combine this is just one example if we can combine the hyperspectral capability of the sensor and the spatial Resolution of the sensor so maybe some of you know this, but maybe Some of you don't know so for nitrogen deficiency the yellowness and the loss of the chlorophyll actually starts somewhere in the middle and the edge of the leaf but for Sorry, let's let's take home for one example the yellowness of the cone Actually start to form when it lose nitrogen but when the cone lose potassium the yellowness start from the edge So if we can combine the spatial resolution and the spectral resolution this gives us the hope of differentiating between them and We are very Lucky that with leaf spec We are probably the first group in the world who can combine hyperspectral and a spatial resolution together So so to move forward for a higher quality of the phenotyping Other questions Thanks You know one thing I've always appreciated about your approach is you're thinking three four five generations ahead on the Technology and what's going to be needed you touched on a couple today? I think that are maybe one or two so I know you're thinking about the next three four or five Anything you can share there about you know what you're thinking the future holds in some of this area That doesn't put you in a difficult spot with respect to ideas today I'm talking about more like a research part So sense of quality is is one part and I think the digital technology to connect people together Is another big part of digital agriculture? So and then we see a lot of the practical values for example now in Indiana State Dicamba pollution dicamba drift is a big issue Yes sensitivity to dicamba is one approach, but some warning signal across the neighborhood The deliver of that warning signal from the government to people Is also important and that can be facilitated by a very mature the digital agriculture data infrastructure So I think the combination of the sensor technology and the data infrastructure combined together can amplify This power and the value so I think the collaboration between the data scientists and the sensor developers This is this is a very critical part for this and I would also say that Maybe the commercial commercialization side how to keep finding new customers who can greatly benefit from this sensor is also important Because that can help us to keep getting the funding from the outside world to facilitate our research So we're not only just talking about farmers because let's take one example the investors in Chicago future market You know these people they make a profit by predicting how much coal and soybean people can produce But you know what they are working on now. They are investing hundreds of dollars To those machine learning algorithm to take in the everyday news and the government policy to try and also the environment and climate change To try to predict three months later. How much grain of the soybean and the corn will be on the ships on the ocean But this is actually too slow too late and we can tell them, okay We phenotyping people we are actually the first to get the data in the field to predict even before the harvesting We already got the data of how much yield people are people will harvest So I think getting the interest of the many different disciplines of people and they involve the whole community into this and keeping Seeking the new applications and the customers can really grow this community and Help everybody to grow together in this. I don't know if I answer your question Yeah, so I hear very interesting story that you have a very nice technology But the farmers may not be paid for them by themselves In the non ag side of engineering. We have a very similar problem So over the industrial 4.0 requirements small and mid-sized companies Cannot afford the transition. They cannot even afford to buy the app. Some of the app is really expensive So even they know that they're facing getting eliminated by the industrial 4.0 They still cannot afford to go through the technology upgrade and we have a very large group of people working on the subject So I think they will be can be very helpful with you as well So afterwards, I'll be happy to connect you with that group people. We'll be happy to thank you Nice talk. Thank you Thank you, I think we have time for one quick question if there's one more I was intrigued by what Jiang you're saying about a field of view Adoption by farmers So your contention, I mean is there research that shows that it is the inaccuracy of the sensors that's driving or What I mean, what are the things in the field of view is pretty much what's available What are the other what are you seeing there at that interface between, you know, what farmers need and is it just is it? just that inaccuracy or other other factors also well We have agriculture experts here That's my my personal view is my understanding of farmers Sometimes I talk with farmers as well. My understanding is farmers. They they have a very practical way of thinking No matter how cool how fancy the technology looks like that's not appealing to farmers You need to show the farmers that in real practice. This technology can really help them to earn money By the end of the year, you know compared with last year when I didn't use this technology and now I use this technology I earn five thousand more dollars this year This is probably the only reason we can persuade the farmers to to invest in today's these things so so that is why I think We and this how to turn the key how to start this this practice is a big challenge So I think this is a role of a university in academia. I think it is our job We should not just try to Advertise and persuade the farmers to use something, but we should do this experiment. We should do this Experimental things and collect data and show the proof that see if you do this you will earn money We already did this now everybody follow us So I think this is a one-row of a land-ground University And this is exactly what we are doing. We have a we have a research firm here near campus 1600 acres and then we have five Purdue research agriculture research centers And we are doing this every year so I think Through our extension team maybe Getting closer to the farmers and by the way Most of the farmers think in one way, but we also have some pioneer farmers And we can also collaborate or of those pioneer farmers who is willing to be the first Be the you know, we call it innovators and frontiers to take this these technologies to to show this future map Yeah Thank you All right. Thank you, Dr. Jen