Fly along for a Rowbot surprise!

The Rowbot platform is ideal for seeding cover crops late summer / early fall into tall corn. Take a look at the results of a fun project inspired by Grant Wyffels.

2018 Rowbot Nitrogen Field Trials

We have several field trials underway in Minnesota, Iowa, and Illinois. Our goal with these innovative trials is to push the envelope on driving farmer profitability by adjusting the timing and rate of nitrogen applications. Look for more on the results after harvest in October / November.

Here's a video of our prototype machine applying nitrogen on a field in north-central Iowa. The machine is traveling at just under 4 mph, and I'm chasing after it with one hand in the air holding my phone to capture video and the other hand holding the throttle down. As the "safety driver" this summer, I logged many miles in this manner--usually with my spare hand keeping corn leaves out of my face.

The self-driving tech is keeping it nicely centered in the row. Note that the white GPS receiver stands at six feet off of the ground.

And another video as I wrapped up a long day treating 30 8-row strips (300 feet long) on each of two fields at our test location in Rosemount, MN.

Rowbot featured in the Star Tribune

Agricultural reporter Tom Meersman wrote up a nice piece on Rowbot in this past Sunday's Star Tribune. The article picked up on our recent participation in Google's Demo Day, and the exciting developments on building our robotics team.

The Rowbot story in four minutes at the 2018 Google Demo Day!

This year, Google for Entrepreneurs brought 11 companies from the Americas to San Francisco to pitch as part of Google Demo Day. Thanks to a nomination from our friends at Work at Coco in Minneapolis, Rowbot made the cut and was able to take part in this fantastic event. See the video below for what the Rowbot story looks like when forced within a four-minute timeline, complete with a loud gong if we lost track of time! Many thanks to Kyle and the team at Fueled Collective, who also provided invaluable pitch help from master storytellers Chris and Jeff with Narrative Pros, Becca and the Google for Entrepreneur's team, the judges for their great questions, all of the amazing people representing the other teams who made it a special experience, and the pitch-whisperer, Ginny Brown for coaxing the best possible performance out of us all.

Reduce total nitrogen applied to corn using multiple in-season (sidedress) applications

At ROWBOT, we are driven to increase farmer profitability by reducing input costs and boosting yields. It has become clear to us that a management plan with multiple in-season nitrogen applications is the ticket to using this costly input efficiently. Our goal is to make these in-season nitrogen applications easy and extremely cost effective.

During the 2016 growing season we carried out two nitrogen strip trials in collaboration with Shannon Gomes (Cedar Basin Crop Consulting), who connected us with a couple of progressive growers in NE Iowa.

Description of the three nitrogen treatments, with in-season applications in mid June and early July.

For the field data shown here, we had three treatments replicated three times in strips that were about 1200 feet long. We are indebted to Clark McGregor, who took care of the early nitrogen applications and harvested the strips.

Treatment A was the farmer's normal treatment--most of the nitrogen was applied during mid June. We added a second in-season nitrogen application for treatments B and C. Based on the Adapt-N model, the recommended application for B and C on July 1st was about 35 lb N/acre. We decided to have the B treatment at 60 lbs/acre and the C treatment a bit lower than the Adapt-N recommendation at 30 lb N/acre.

This is how the total nitrogen in each treatment stacked up. The "Reference" is included to show how much the grower would have applied if all of the nitrogen were applied at or before the time of planting.

The harvest data are in, and the results are quite impressive! Even with 15 lb N/acre less in treatment B, the measured yield was slightly higher than the control--treatment A (statistically, there's not a difference between the yield of treatments A and B given we have just one set of strip trials). Surprisingly, treatment C that had a whopping 45 lb N/acre less than the control had only a slightly lower yield (the difference is only marginally significant).

Yield data from the three nitrogen treatments. Grain from each strip was measured in a weigh wagon and moisture estimated from the combine's yield monitor.

The addition of a second in-season nitrogen application meant that we gained about two weeks of additional insight into the season before the final nitrogen application was needed. Plus, not all of our nitrogen was on the field during late June, which was a fairly wet period (see comparison to other seasons below).

We are making plans for an expanded set of field trials for the 2017 growing season that should further underscore the value growers can realize through multiple, in-season nitrogen applications using the ROWBOT platform.

ROWBOT featured in a Beck's PFR Report

Earlier in 2016, the team behind Beck's Practical Farm Research (PFR) reached out asking if we could take part in a seeding trial. That led us to the outskirts of Indianapolis during early September to lay down cover crop seed using the ROWBOT platform.

As is outlined in the great video (below) produced by the PFR team at Beck's, they wanted to compare results from a simulated aerial application, seeding with a drill after corn harvest, and a high-quality broadcast seeding by our ROWBOT platform.

They'll be collecting data and imagery from the trial, which we'll post here. Sounds like the stand resulting from the ROWBOT is quite uniform compared to the simulated aerial seeding. As expected, the drilled seeding is very uniform, but its growth was delayed by several weeks given that they had to wait to do that seeding until after the corn was harvested.

Seeding a cover crop on an organic corn field with the ROWBOT platform

Last week, I did some cover crop seeding at the Fitzgerald farm in central Minnesota. Those are young red clover plants that were seeded on Sept. 19th, photographed yesterday, less than two weeks later. If you are used to looking at photos on our site, you'll notice some vegetation (a.k.a. weeds) on the ground between corn rows, which is typical in an organic operation (incidentally, removing weeds on large organic fields is a big opportunity, which is a good match for robotics).

Yesterday, I was able to grab a couple of videos while chasing the machine at 4 mph (a very fast walk on the relatively uneven terrain of a typical field). Enjoy!

Laying down cover crop seed the ROWBOT way

Just back from a few hours of seeding cover crops at our test site on the University of Minnesota's Research and Outreach Station in Rosemount, MN. Once again, we're collaborating with Prof. Scotty Wells, who has a number of other cover crop studies underway. This is also part of the New Agricultural Bioeconomy project at the U of MN, which has received funding from the U's MnDRIVE initiative.

With this study, we are testing the impact of planting date on cover crop biomass. Why? Well, it's tough to get a successful cover crop following corn if one waits until after harvest, which may occur just before the snow starts to fly. Being able to get the cover crop established at the end of the summer or early fall helps it get established well before the onset of winter. The ROWBOT way, so to speak, is to get in when corn is mature and lay down cover crop in a best-of-class broadcast application. This study is designed to confirm what is logical: an earlier planting date should yield a more robust cover crop with more biomass, which is a decent proxy for the soil health benefits provided by cover crops.

We planted a round of plots today (9/15), and another set of plots two weeks ago (9/1). There will be one more planting date around the middle of October.  Last year's initial study, while not as rigorous as this year's, suggested that less biomass results from later planting dates. That makes sense, and it is the reason we're working hard to commercialize the ROWBOT technology so that we can seed cover crops on millions of acres long before corn is harvested!

Check out the photos from today of the cover crops seeded two weeks ago. Note that there were a few rain events after planting (over one inch on 9/6 and another half inch on 9/7).

Here you can see the cereal rye seeds photographed just after planting today.

Cover crop seeding results on Earth Day

This past fall we reported on our cover crop seeding trials done in collaboration with Prof. Scotty Wells at the University of Minnesota. Below are a couple of Earth Day photos taken of two of the plots. We used the ROWBOT platform to plant both plots with cereal rye--a fast growing species that is a trusted cover crop because it is easy to get established, especially later in the season.

We'll be getting out soon to do a careful biomass harvest, however, the qualitative result is that there was more growth on the plots that were planted earlier in the fall. If this preliminary result holds, then it will underscores the value of getting cover crops seeded well before corn harvest.

During the 2016 season, we plan to repeat and expand these seeding trials. It'll be important to get some seeding dates in August, too. We believe that seeding cover crops into tall corn using the ROWBOT platform during August and early September will provide excellent results, even in the northern part of the Corn Belt. Stay tuned!

ROWBOT featured on This Week in Startups

I was recently interviewed by Jason Calacanis on his show, This Week in Startups. It was a great opportunity to discuss our view of the future, which is that small, autonomous machines will play a large role in carrying out in-season work for crops like corn. Plus, these small platforms will be key for collecting actionable data on farms, whether to improve the job to be done that day, or to help farmers make better decisions later that season and beyond.

The real magic of small machines designed to work throughout the growing season (think fleets of small, between-row machines that operate regardless of crop height) is that farmers will have repeated opportunities during the growing season to adapt their management to actual growing conditions that season.

While the Buffetts (Howard G. and Howard W.) are spot on in arguing that farmers have historically had 40 Chances to produce a great crop, we believe that farmers of the future should have many chances to improve a crop within a given growing season. We expect iterative improvements in other industries--why not in farming? ROWBOT has the platform that can deliver on this promise.