Drones, Sensors, AI: Farming the Future with Dr. Bai

 Dr. Frank Bai shares his global journey into precision and digital agriculture and highlights how sensors, drones, and AI-driven tools are shaping the future of farming. As an Assistant Professor at NC State, he bridges engineering and agriculture to tackle challenges in the field. In this episode, he discusses his autonomous scouting project at the Sandhills Research Station, innovative methods for screening drought tolerance in soybeans, and the growing role of digital technologies in making farming more efficient, sustainable, and affordable. The conversation also explores the importance of collaboration, technology adoption, and equipping the next generation of farmers for an increasingly automated, data-driven future. 

Transcript

AgTech360 discusses breakthrough technologies that are impacting growers, businesses, and consumers. Hear from industry and academic experts about what's on the horizon.

Welcome to AgTech360 and our special series, Harvesting Innovation. From drones, flying over fields to sensors, capturing every detail of soil and weather conditions, digital tools are transforming how we grow crops. Today's guest, Dr. Frank Bai from NC State is at the forefront of this work. With expertise and sensor-driven decision-making and precision agriculture, Frank is helping to shape more sustainable and affordable farming systems for the future.

So thanks for joining me, Frank. Let's get to it.

Sure. Thank you very much, Adrian. Thanks for having me.

Yes. So let's delve in a little bit into your background first. I understand you've got some degrees from a few different countries, including China, Japan, and now the US and how did all those different experiences lead you to NC State? So tell us a little bit about your background and what drew you into this field.

I think I have a unique experience, since you mentioned I got degrees in three countries. So when I was young, I have no idea. I will find a job at NC State to do research and teaching in the area of precision and digital agriculture. But when I looking backwards, actually I can connect these dots very well. All the things started from my undergrad degree. I'm doing hydraulic and hydropower engineering, so I have background in mechanics and hydrology. I was pretty good at it. But then I found out for my senior design project, there is a very interesting project in ag engineering. So at that time, a quite famous professor in our department developed a computer vision-based method to quantify the soil infiltration rate, the water infiltration rate at different soil textures. So that one technology looks very promising and it's labor receiving. So basically I just did my senior design project in ag engineer.

So then I went on my master degree in irrigation system study, and then I went to Japan for my PhD to study the drift reduction. And after that I got a job offer from Lincoln, Nebraska. But since then I started to know more about sensing. So from last year, July 2024, I joined NC State as assistant professor in BAE Department, which is biological and agricultural engineering, and now I think I can leverage what I have and doing teaching and research in precision and digital agriculture.

You seem to take a very systematic approach to agriculture from scouting through sensing to decision-making. Can you kind of walk us through how that whole process works in practice?

Since I'm also teaching precision ag technology courses, so we always teach in a systemic way. So basically if you think about any fieldwork in agriculture operation, every year we just do a cycle. So from sensing, maybe planting and cultivation all the way to harvesting. So the sensing is happening everywhere, it happening in the machine, it happening in the air, it also happening on the ground. So basically this sensing technology will help farmers or growers in every steps in this cycle to improve the farming efficiency.

So I know one of your projects is you have a large research trial at the Sandhills Research Station here in North Carolina on soybeans, I believe. So can you tell us a little bit about that project and what you are hoping to learn from it?

This is quite exciting project and it's a going on project. I will try to introduce this in a good way. So basically I have collaboration with Dr. Ben Fallon and I got strong support from the Station as well. So basically we are doing... The first objective of this research is autonomous scouting. So think about one day you can have autonomous drone system can fly out to the field anytime you want every morning or one day before it will send you a report like what decision you need to make the next day or today to do some maybe site specific field management to minimize your resource input and also boost your yield.

So we are doing autonomous drone project as one of the main goal. And yesterday we just did the first fully remote flights. So my colleague Justin was controlling the drone from Raleigh. And I believe we're the first group to be able to do this in the CALS College. The second goal for myself is to try to help our collaborator, basically the soybean breeder, to try to rank the drought tolerant capacity from hundreds of genetic lines in that experiment. Last year we published a paper to try to integrate the phenotyping, the crop modeling and also soil sensing to try to provide a high-resolution phenotyping tool to directly modeling the crop water use.

How do you see drones, sensors and other digital tools helping us better understand weather, soil and crop conditions and helping farmers make good decisions?

Since I'm teaching precision ag technologies, so we always thinking three parts, the sensing, decision making and implementation. If the sensing technology can be well integrated into this scenario, so the whole workflow can work out by itself not only in the sensing but how to convert the sensing into action. If we can also automate this process for most of the operation, the sensing technology will benefit the whole farm operation a lot.

How does this then translate into helping farmers be more efficient and sustainable with their farming practices?

One of my main goal in the long term is really try to develop some affordable and high-performance digital tool to help farmers to reduce the labor cost, improve the resource, use efficiency, and also have great yield. So with that three objectives in mind, if we can develop the digital tool, the farmers can really use their technology daily or frequently.

The North Carolina Plant Sciences Initiative impacts lives through innovative applications and discoveries. By leveraging cutting-edge research and technology, we address global challenges related to agriculture, sustainability, and human health.

So with that in mind, and looking forward into the future, I guess with your crystal ball, how do you see these technologies evolving and helping farmers in say 5 or 10 years? What will be different then from perhaps what we're seeing today?

Most of us realize how artificial intelligence is changing every sector of our industry. And I think that also include precision ag. Basically we already have fantastic sensing technology at different scale on different sensing platform. If you combine the different sensors with different sensing platform, actually it cover most of the aspects of the data connection from the ground to the satellite. And with the help with big data set and AI, the more and more powerful model can be built, which means more accurate, more scalable models to predict and help us to make the decision better and also predict the yield and other abiotic stress or biotic stress. So I would say in the next 10 years, more automation will happen with the better integration of AI modeling the big data and ag machines.

So let's switch gears and talk a little bit about collaboration and industry partnerships. I know you do a lot of both. I know one of your collaborations is with Professor Chris Reberg-Horton, who's been on this podcast in the past and you're working with him and other colleagues in your biological engineering department. So how do you approach collaborative projects like this and what do you look for in industry partnerships?

I think the first step is to have a great idea for collaboration. And for me, since I have been doing crop sensing for a long time, sometime I can have some great idea for collaboration and then I will try to think about how this idea can benefit the entire team. With that in mind, I think a great communication channel will be very helpful. So basically as a early career researcher, I really hope all us can have a very supportive environment when we do collaborations. Luckily, I believe I'm in a very good collaborative environment with my collaborators and that is really good for me.

For the second question about industry partner, I think our university or college and PSI can play a very important role to partner with industry players to develop or to further develop our technologies into products.

So now let's talk a little bit about adoption, these new tools, I think in our everyday lives, digitalization has taken immense steps in almost every aspect that we can think about, but maybe not so much, at least until now in agriculture. Tell me a little bit about, do you think that today's digital tools that are being offered to farmers, are they user-friendly? Do they need to be improved from that perspective?

Speaking of the adoption rate of different agriculture technologies, especially the precision ag technologies, the first reaction is there are excellent technologies which are widely adopted and what will be the main reason. So take auto-steering for example, with advance of the GPS technology and most of the new tractors have the auto-steering function, which can significantly reduce the working load from the drivers. If one technology can be user-friendly and can be used by farmers in every day's operation, I think that will be a success.

And the other aspect that I heard you talk about then was a return on that investment. Some of these technologies can be expensive and I think is obvious that a farmer wants to see a return on that investment.

That might be one big challenge. If you think about in North Carolina we have growers have different scales and most of the technology if you calculate the return over investment.

With all that being said, what new skill sets do farmers and operators need to have to use these new digital ag tools? And are you seeing with farmers that are currently adopting or trying to adopt these tools a difference in say, generational effects?

I would say the first skill required will be the knowledge. So they need to basically understand how it works and they need to learn the knowledge base. And in the school we teach new generation of farmers all the new cutting-edge digital tools in the class, that will be a plus in their education experience.

And last question from me. As someone who's working on the cutting edge of research, what do you think the farm of the future is going to look like in 20 years?

I think in 20 years that will be very different. So first of all, we still need machines in the field, that won't change, but how it works. The one short answer is more automation, with the help with AI, robotics and modeling, we should be able to integrate all this software and modeling into existing machines and there are existing autonomous tractors on the market. With a new business model, I think there will be more automation in the field for most of the operation in the field crop management side, and there will be less labor requirement. And also we need farmers equipped with more digital skill.

Thanks for sharing your time and research expertise with us today. It's really inspiring to see how digital tools and sensor technologies and automation are going to be paving the way for more sustainable and efficient farming. Appreciate you joining us and thanks for being on this Harvest Innovation podcast series.

AgTech360 shares relevant news and breakthroughs with audiences across the globe. Stay connected and join the conversation by following NCPSI on social media.