Upstream Ag Insights - August 8th 2021

Essential news and analysis for agribusiness leaders

Shane Thomas

Aug 8, 2021

Welcome to the 79th Edition of Upstream Ag Insights!

Index for the week:

Autonomous Farming: The Future Comes Faster than You Think
John Deere Acquires Autonomous Company Bearflag Robotics
FBN Extends Trialling Capabilities
Curing What Ails Agronomic Efficiency
Drone Usage Accelerated During COVID-19 Pandemic
Deveron Announces Public Placement Financing
Trimble Launches Venture Fund

Have a great week!

Join thousands of agribusiness leaders and get essential news and analysis right to your inbox every Sunday!

Autonomous Farming: The Future Comes Faster than You Think

One of the hottest topics in agriculture is autonomy. This is exciting to me for several reasons: the ability to decrease labour constraints, removing inefficient gut feel from decisions and practices and the potential for new business models. And if we begin to break it down, autonomy is here today, on almost every farm and agribusiness!

From automated product control, to auto boom adjustments, to auto scripts to auto steer. This is all a form of automation.

At the end of July I was invited down to the Raven IGNITE event in Sioux Falls, South Dakota. Raven is in a compelling position when it comes to autonomy having acquired DOT, Smart Ag, Jaybridge Robotics and AgSync. This event really solidified my view that the future of autonomy is already here, it just isn’t evenly distributed.

It was especially evident while watching the autonomous OMNiPOWER unit move up and down the field:

As cool as watching the autonomous unit go through the field was, the nerd in me was more enthralled by one specific image shown that I think encapsulates a guideline for autonomous evolution, consistent with my comment that “autonomy is already here”:

Source: Raven Industries

While it’s easy to get caught up in the semantics of what indicates a level 2 vs. a level 3 the reason I latched on to it was the fact that it frames autonomy as an evolving journey, unique to each operation. Not a monolithic end point.

If we look at Ravens aforementioned acquisitions and core business we can see their acquisitions align specifically with this journey: core competency in software controls, acquiring AgSync to augment dispatching and equipment positioning, followed by SmartAg to enable automated tractor driving and then DOT, a fully autonomous platform. This all being augmented by the JayBridge IP acquisition that includes technology related to path-planning and obstacle detection and avoidance. Not to mention autonomous precision applications in their strategic alliance with AGCO and xarvio/Bosch for targeted spraying technology.

You can’t have full farm autonomy without having these various layers

Raven has capabilities at every level of autonomy. The capabilities lay out an entire autonomous tech stack for farmers to retails with equipment fleets.

Business Models

In the autonomy frame of mind, on the flight back home I listened to this AgTech So What Podcast: Is the future of farming hands-free?

What I always enjoy about the AgTech So What Podcast is that Sarah Nolet always does get to the “so what?”, in this instance specifically, the potential business model evolution and not just the fact that there are tractors without humans:

the evolution of the autonomous agriculture landscape over the past decade continues to see more startups than incumbents, given that autonomy tends to cannibalize the dealership-heavy business model of existing agricultural equipment companies.

Companies like John Deere have stuck with traditional approaches to their business: sell more and higher value (more up sells) tractors + generate service revenue. It’s even apparent in their 2020 investor presentation based on their forecast of margin growth being associated with this traditional approach to selling:

I have touched on unique equipment business models in the past taking a Rolls Royce approach and suggesting it might make sense in agriculture which flips the script on business models:

Today Deere sells a tractor one time and can theoretically benefit if that piece equipment doesn’t last long and needs more service/parts. What is possible is a model that is similar to what Rolls Royce does with airplane engines - charging airlines on a Power-by-the-Hour basis while continuously monitoring the engines which de-risks the airline up front and throughout the life of the engine, proactively preventing unanticipated maintenance costs. On top of this RR gets an enhanced relationship with the customer, customization capabilities for the customer and recurring revenues. Now think of the airline as a farmer.

Pay for use as it is used versus up front costs. This better aligns incentives for the farms or agribusinesses with the performance from John Deere.

Given different business models, there are still diverging approaches to autonomy:

Fully autonomous robots vs. hardware/software packages that are retrofit to existing equipment
Autonomous vehicles designed to perform specialized tasks (e.g., harvesting, spraying, etc.) vs. the autonomous equivalent of the utility tractor
Companies developing the “full-stack” required for autonomous vehicles vs. a marketplace for componentry (e.g., arms, end effectors, mapping and analysis capabilities, etc.)

This can even be swarms vs. larger equipment. I think it’s worth noting the autonomy extends beyond tractors and out to other aspects too, like drones.

The more I have thought on this the more I like the “full stack” development by a company (granted this doesn’t mean it’s independent of other aspects listed in the bullets above).

End to end integration and execution with a concept as complex as fully autonomous (controls to equipment operation aka level 5) equipment is the approach successful companies will be focused on and this will likely be a driver of continued consolidation in the autonomy space.

Challenges

In the past I have gotten hung up on the challenges inherent within autonomy: bylaws for autonomous equipment transportation, pushback from personnel because it’s “stealing jobs”, costs or a bias for traditional approaches just to name a few.

But there are big dollars and big organizations working to alleviate these concerns, not to mention the opportunity for redeploying idle labour to higher ROI roles (or eliminating the jobs people hate that many in the autonomy space refer to as “dull, dirty or dangerous”) and the liklihood of higher value job creation - after all, as humans we have been demonizing technology for it’s elimination of jobs since the industrial revolution and we all still have jobs thanks to innovation driving specialization and the division of labour.

Even the comment I hear frequently is this: “but I enjoy my time running machinery”. People said similar things about horses in the 1890’s. And there will still be opportunity to derive some of the field level knowledge operators derive while having autonomous equipment…autonomy doesn’t have to mean the cab will be removed entirely from every piece of equipment.

The thing to keep in mind: autonomy is a journey, it is going to be one aspect of running equipment, like auto steer, then an autonomous turn capability…eventually eating up all small tasks of running equipment until there is a natural evolution to “level 5” or fully autonomous farming. We may not choose full autonomy today, but overtime we will see choices towards incremental autonomous capabilities that leads us towards full autonomy. Autonomy is forecast to grow at a rapid clip by organizations like John Deere (see next story), CNHI and independent reports like this one suggesting a >30% CAGR.

Change happens slowly, then all at once. Therefore it’s worth considering today as agribusiness professionals:

How would a farm running autonomous equipment impact your day to day?
How does it impact the value you provide to the farmer?
What does autonomy mean for the products you sell, support or service?

I wrote the above on Wednesday night and on Thursday the below John Deere announcement came out further reinforcing the view towards autonomous future coming faster than we think:

John Deere Acquires Bear Flag Robotics to Accelerate Autonomous Technology on the Farm - Newswire

Deere & Company has signed a definitive agreement to acquire Bear Flag Robotics for $250 million USD. Founded in 2017, the Silicon Valley-based startup develops autonomous driving technology compatible with existing machines. The deal accelerates the development and delivery of automation and autonomy on the farm and supports John Deere's long-term strategy to create smarter machines with advanced technology to support individual customer needs.

This acquisition comes after a strategic alliance between the two companies over the course of the last ~1 year.

The value of this acquisition is large…$250 million with a total amount of capital raised coming in at only $8 million according to Crunchbase.

For context, Raven industries acquired DOT Autonomous Farming and SmartAg for a total of $72 million USD. CNHI is now acquiring Raven Industries for $2.1 billion, however, Raven comes with numerous established and profitable businesses so it is impossible to compare the acquisitions outright.

What is note worthy as well (although Deere could change this as the focus is likely more on the technology itself I suspect in the case of this acquisition) is the business model of Bearflag Robotics.

From their website:

Bear Flag provides farms with complete autonomous tractors charging farms only for the work that is done

Charged by use. You don’t buy a Bearflag tractor, you are charged on a per acre usage, similar to how custom services would be today. This is different, but similar to what was alluded to when I brought up the Rolls Royce model in airplanes. Could this be an approach by John Deere in the future?

My last comment is around the focus of BearFlag: tillage passes. Some of the hesitation around autonomy is a screw up of a high value activity (eg: seeding), costing significant dollars. We all have trust issues with technology, I’ve highlighted my own automatic email send hesitation when I started Upstream even.

Not only is tilling low risk, it’s also boring. A perfect starting point.

What is noteworthy too is the focus. They didn’t attempt to do autonomous passes across numerous areas, they focused in on one specific pass that has the aforementioned tendencies which makes it a smart starting point in my opinion.

Big news overall. CNHI and Deere making continued plays in automation.

What does an organization like AGCO do in response to CNHI and John Deere? There are a number of organizations out there with capabilities in the realm of autonomy and operational efficiency; from Solinftec (AGCO currently has a partnership with them), to Blue-White Robotics to Verge Agro.

One more related note: if we look at the SEC proxy filing from the CNHI acquisition of Raven Industries, there were two other bidders on top of CNHI engaged during the 6 month process. It doesn’t state who, but if I was betting I would think one of them was AGCO.

Sticking with AGCO:

Precision Planting Agrees to Acquire Headsight Business - Precision Ag Dealer

Precision Planting, LLC a subsidiary of AGCO Corporation, a provider of precision ag technologies to help make farmers smarter every season, announced an agreement to acquire the business and assets of Headsight, Inc. (Headsight), a precision agriculture harvesting solution company.

The interest to me here is “precision agriculture harvesting”. We talk about precisely applying fertilizer, or positioning seed in the soil. All aspects to manage costs or increase yield. What isn’t talked about is the precision necessary to capture that yield aka minimize harvest losses at the combine.

This is where new technology has an opportunity, specifically AI vision technology to more precisely assess losses and calibrate the combine are a big opportunity, like Farmwave. This is a huge opportunity and one I think we will see companies continue to innovate around.

Share Upstream Ag Insights

This weeks edition of Upstream Ag Insights was brought to you in partnership with GeoPard!

GeoPard Agriculture is a cloud-based powerhouse for precisionAg data analysis, prescriptions creation, smart scouting, and soil sampling. GeoPard Variable Rate (VR) prescription maps are used for fertilizing, crop protection, seeding, irrigation, desiccation, optimization of crop inputs, and yield while preserving resources. The GeoPard engine processes satellite imagery, soil sampling, yield, high-density soil scanners (EC/Moisture/Soil compaction), and topography data, building sophisticated multi-layer Field Potential maps and 3d models. GeoPard VR maps are compatible with most existing machinery and terminal brands. Moreover, there is a bi-directional seamless integration with the MyJohnDeere Ops center, which enables John Deere clients to receive GeoPard Rx maps, crop monitoring, and topography analytics wirelessly.

GeoPard produces precisionAg analytics at scale and provides it via API, widgets, web, mobile applications, and White label solutions for crop advisers, ag input companies, dealerships.

The GeoPard team has great experience of launching successful precisionAg software products since 2012 (Zoner.ag (powerZones) acquired by Bayer, then Xarvio).

FBN Opens 2022 Applications For On-Farm R&D Trials - Successful Farm

Farmers Business Network (FBN) has begun taking 2022 applications for the FBN On-Farm R&D Trial Program, a unique testing program for sustainable agriculture and innovative technologies that connects developers of agricultural technology solutions – from biologicals to robots and sensors – directly with farmers and large-scale, on-farm, real-world testing environments.

Strong on farm testing is a benefit to farmers. I have commended FBN for their R&D efforts in biologicals in the past and suggested retailers have an opportunity to lead in the bio trialling space as well.

What’s most interesting about this announcement though is the expansion of the trialling program and similarity to another company in the farm trialling business: IN10T, a digitally-powered, data-driven company focused on accelerating the adoption of emerging agricultural innovations. They do so by partnering with farmers to test and learn the benefits of innovative new technologies and products in real farm settings.

In March I stated this about IN10T:

One of the start-up companies I am most bullish on is IN10T.

IN10T has software and tools that power insights about products, practices and systems. They start with the farmer experience and data in mind.

They define themselves as doing the following:

platform service
product performance
farmer experience
commercialization strategies

The last aspect is a commercialization strategy. This is where IN10T would support the company and their go-to-market strategy, such as determining an approach to distribution and accessing the market and product positioning.

The commercialization aspect is noteworthy because in the case of FBN, they would offer that route to market.

This means in part what FBN is doing is a backwards integration and doubling down on these trialling initiatives that feeds three areas of their business:

Which products to bring into their marketplace (product expansion)
Sales Support Data (and decision support for farmers)
Identifying Products for Other Program Initiatives (eg: Gradable programs).

Partnering with FBN for many input companies, as of today, means that if a company opts to work with FBN, they likely shut off the traditional channel to market.

What differentiates the IN10T approach to FBN is at minimum two fold:

Digital systems - Not all software platforms work in all settings and IN10T’s is built with a focus on being effective for trialling needs. This includes dash boarding, reports and specific functionality.
Staff focused on the trialling aspect - A specialized and focused approach to working with the farmer is important to successfully execute the trial.
Trials are simple in nature, yet complex to do effectively. That’s why those above two aspects are so important as differentiators.
It’s important to state that I think there is room for both. But this approach reminds of a physical world “embrace and extend” strategy, that Microsoft so famously executed and was criticized for by the US government for being anti competitive (Note: I am not suggesting FBN is being anti competitive). Using the trialling as a support and extension of their current offering.

What is this strategy?

The strategy's three phases are:

Embrace: Development of software substantially compatible with a competing product, or implementing a public standard.
Extend: Addition and promotion of features not supported by the competing product or part of the standard, creating interoperability problems for customers who try to use the "simple" standard.
Extinguish: When extensions become a de facto standard because of their dominant market share, they marginalize competitors that do not or cannot support the new extensions.

The embrace and extend aspects are primarily a digital phenomenon and they have began to play out in the social media space, albeit to a lesser degree. Instagram (Facebook) famously took the “stories” concept from Snapchat and embedded it within their own platform. Stories is where an individual can post a video/image publicly for their followers, friends or anyone to see for a 24 hour time frame. This proved to be an extremely sticky and compelling way for young people to share what was happening in their life. With Stories, Instagram wasn’t embracing a standard so much as a medium.

While the FBN approach is different because of the physical nature, what FBN is doing is embracing a particular trialling approach (trials-as-a-service) and connecting that with their unique distribution ability (today ~70 million acres) aka “extending” the value proposition to their company trial partners.

This distribution is relatively small and the market for great trialling and product insight is large (plus, there are no interoperability needs) which is why I wouldn’t consider FBN’s efforts in this space to equate to an “extinguish” of IN10’s capabilities elsewhere.

Additionally, Microsoft was a true platform that had a monopolistic market share of computers that empowered them to execute this. Today, FBN is not a platform and is still adding capabilities without one “second to none” value proposition.

What FBN wants to do though is leverage trialling into their ecosystem because it can feed other areas of their business as mentioned earlier. They likely view “trialling” as a feature and intensifier of their flywheel. While trialing may seem like a distracting initiative, if this gets executed on effectively it should be synergistic to their business.

Perspective: Curing What Ails Agronomic Efficiency - Precision Ag

This is a really well done article with some great perspective. A lot is consistent in how I think about the future of the trusted advisor, albeit I get there in a slightly different way.

Some notable quotes:

Ag retailers frequently assert that service is central to their competitive advantage, which is realized as prompt service, expert advice, and just overall resilience to get things done. However, all these differentiators (in their current state) have one thing in common: the reliance on humans to deliver on them. I know this may come across as harsh, and maybe even callous. But in comparison to many of the technologies available to us today, humans are inefficient, inconsistent, unreliable, and have an inevitable focus towards self-interest.

I agree. In fact I don’t think it’s harsh or callous, I think the biggest flaw we see in humans no matter the industry is a failure to acknowledge this reality.

The need for human-verified (agronomy decisions and info) is a common denominator, and this is an impediment to scale and efficiency.

One can argue there doesn’t necessarily need to be a scale benefit, but the reality is a lot of the promises of technology are the opportunities for scale.

Lets consider Farmers Edge* who went public positioning themselves as a software-as-a-service business, but they aren’t. They are a physical service business supported with software. Physical services have marginal costs, time and efficiency constraints, and limitations on reaching “escape velocity” that true SaaS business don’t…Unless they manage to eliminate the need for people consistently across their services while keeping an ROI for the farmer.

Companies need to continue to look for incremental automation to manage costs as well as better outcomes for farmers removed from human emotion and bias.

Lets break it down further: Every job is made up of a number of responsibilities which are made up of numerous tasks and then micro-actions once broken down to the smallest possible components.

An “agronomist” is a job, that has responsibilities (eg: sell product, support relationships, solve problems etc) that are executed via tasks (eg: scout field, make fertilizer and chemistry recommendations etc) and are made up of actions (eg: drive to field to acquire information to inform a recommendation).

In a softer skill and knowledge based job like agronomist it doesn’t just get “automated”. Small actions get eaten up by technological capabilities until eventually the “job” itself can be fully automated. This approach is what eventually helped me come to a similar conclusion to the author.

Thinking about it like this allows us to ask:

What actions can be automated/replaced remotely or otherwise today?
What is the frequency of the actions?
Is the automation as effective as the human?
What is the timeline to automate other actions?
What tasks and actions can’t be automated? There will be some aspects that simply cannot be eliminated.
What new tasks and responsibilities are created as this plays out?

Today there is still a requirement for a human across most responsibilities, but there will be continued progression towards automated aspects. As automation progresses, there will be a shift towards new products/services, systems and business models necessary. This means how agribusinesses define “service” and “value added” will continue to evolve.

The efficiencies realized by eliminating the “human-in-the-middle” are undeniable. It takes courage to accept a transformational vision and the willingness to invest in technology that turns human assets into labor-free processes. However, those who take this path will gain a competitive advantage, scale to succeed and leave a legacy for generations.

As the author eloquently states, it takes vision and courage to acknowledge and invest in this change.

*Disclosure: I was employed by Farmers Edge in 2019 and 2020.

Related: Technology Is… Considering A CTO - AgWeb

I like the message of this article, whether focused on a farm or an agribusiness itself. Companies do need a CTO that has the core information system competencies, and they also need an implementation champion(s). They are different, but both important.

Share Upstream Ag Insights

Deveron Announces Private Placement Financing - Yahoo Finance

Deveron, an agriculture digital services and analytics provider, is pleased to announce a non-brokered private placement financing for gross proceeds of $3,500,000 through the issuance of 5,384,615 units in the capital of the Company. Gross proceeds raised from the Offering will be used for working capital and general corporate purposes.

This is notable because Deveron* is currently executing a roll up strategy unlike anyone else in the industry: acquiring technology plus agronomic consulting companies that enable distribution to deploy the tech.

In the last year they have acquired either directly or via joint venture somewhere around 5 different companies including technology based, agronomic service based and soil health/soil lab based. The majority of this was after their December 2020 private placement financing, so this announcement seems indicative of a reload. Therefore, I am expecting more acquisitions from them this year and into 2022.

Their acquisitions have focused on acquiring boots on the ground agronomy consulting companies and I would suspect this round of financing will continue to focus in on that continued acquisition of agronomy companies that give them cash flow and access to more acres for their labs and technology.

*Disclosure: I am a shareholder within Deveron (TSX.V: FARM)

Ag, Programmatic, and the Human Element - WS

They attacked it from the standpoint of marketing in agriculture, a niche B2B vertical, focusing on the value of programmatic to an industry that typically shies away from it, the value in hyper-targeting, and the human element of programmatic.

Drone Usage Accelerated During COVID-19 Pandemic - Future Farming

Q2 2021 flights and users nearly doubled (+99.7% for flights, +82% for users) from the same quarter last year.

I have talked about my new found bullishness for drones since the start of Upstream for those wanting to dive more into some of the factors driving the increased usage.

Trimble Launches $200 Million Venture Fund - Trimble

Trimble announced today the creation of Trimble Ventures, a venture fund focused on investing in early and growth-stage innovative companies that align with Trimble's mission of transforming work in the agriculture, construction, geospatial and transportation industries. Trimble Ventures will deploy strategic capital to accelerate the growth of innovative companies and partners that complement Trimble's products, technology platforms and support its customer's work. The fund will invest in early and growth-stage companies with technologies and solutions related to hardware and software applications; artificial intelligence; augmented, virtual and mixed reality; autonomy and robotics; blockchain; the Internet of Things (IoT) and analytics; and sustainability.

Tech Can't Make It Rain - Prime Future

I loved this very practical perspective from Janette on the realities and limitations of technology in ag.

This perspective brought up a concept I’d read about several months ago, specifically what China is doing: cloud seeding. We see cloud seeding used to today by insurance companies to manage hail damage. But China is literally trying to manipulate where and when it rains!

Last month, 16 “artificial rain enhancement rockets” were launched off the back of a pickup truck 300 miles south of Beijing. The operation, ordered up by the Juye County Meteorological Bureau in response to a local drought, was reportedly a success. Over the next 24 hours, the county received more than two inches of rain that, according to local officials, alleviated the drought, lowered the risk of forest fires and improved air quality.

This may not be agtech, but there is definitely an aspiration to “make it rain” via technological influence. The bigger implications of this are obviously complex, but that’s a different conversation!

Non Ag Article

Compounding Crazy - Not Boring

Sharing this specific non-ag article was inspired by my interest in autonomy over the last couple weeks. Like I mentioned, the future comes faster than you think:

Things that seem anomalous and futuristic today will seem quaint and antiquated in a decade. Today’s newsworthy events will become common occurrences. The things that 2031’s naysayers naysay will be practically unimaginable from today’s perspective. We evolved to look to our past experiences in order to anticipate (and survive) future events. But as the pace of progress increases, our past experiences have a shorter shelf-life of usefulness.

Where does some of our preference for the past stem from?

You have to remember something about what it’s like to stand on a time graph: you can’t see what’s to your right.” It’s possible to look backwards with relative certainty; it’s impossible to look forward with any certainty, particularly when progress itself is increasing at an increasing pace.

One of my favourite sayings is that we stand on the shoulders of giants. We are constantly progressing because of research, experiences and leaders from the past, the capital expended and the convergence of various disciplines. This makes human knowledge compound and the thing about compounding is that it turns into exponential growth of data and innovation:

This also stems from trends of increased platforms (eg: internet based, or companies like Gingko Bioworks) and collaboration.

The challenge is that we are running “caveman software” in an “exponential world”. Exponential growth is difficult for our brains to comprehend.

It’s good to be grounded in the reality of today, but there is a need to keep an eye on the future.

Upstream Ag Insights