Listen to this article
Electric Sheep, a provider of autonomous large-scale outdoor maintenance, is acquiring two commercial landscaping businesses, Phenix Landscape and Complete Landscaping. The acquisitions are part of the company’s long-term strategy to consolidate the landscaping industry.
Tennessee-based Phenix Landscape and California-based Complete Landscaping are both traditional outdoor service providers. Electric Sheep’s plan is to progressively transform these companies’ operations to include its autonomous technology.
The company sees its strategy to roll up traditional landscaping companies as a profitable way to acquire data. Electric Sheep can use this data in a reinforcement learning operational sandbox, where it bakes the foundation model its building and can apply that data to its robots, according to the company’s co-founder and CEO Naganand Murty.
“Our intent is to roll this up to hundreds of millions, to almost a billion dollars, in revenue and deploy at the scale of tens of thousands of robots because we believe that the next shift in robotics will have to happen with this progressive automation and reinforcement learning done up to that scale,” Murty said.
Electric Sheep’s acquisition strategy
There are a few criteria that the company takes into consideration when picking landscaping companies to acquire. The first is that it’s looking for a diverse set of data, which can include a diversity of workflows and sites. According to the company’s co-founder and COO Jarrett Herold, Electric Sheep takes other factors into consideration as well.
“Culture is a really big one, making sure that they have the right people in place and the right book of business and customers, and that they are also running a well-run business, obviously,” Herold said.
The landscaping industry, as a whole, is undergoing a period of consolidation. According to Herold, the typical landscaping business owner is someone who started a small business in high school or college and then continued to grow it up to a business worth millions of dollars. These owners typically reach a point where they’re interested in taking a step back from the business, which is when they start considering selling their business.
It’s also an industry that is well suited to robotics. Landscapers are typically working with low margins and high labor costs, leaving a lot of room for robots to reduce the industry’s labor challenges.
“We can take what’s one of their biggest labor cost centers, mowing, and begin to eat away at that with our robots,” Herold said. “Just showing that vision and how we think about shaping the industry has really sold a lot of landscapers.”
Once a company is acquired by Electric Sheep, the team starts the process of gradually introducing robots into existing workflows.
“We acquire and then we remap workflows because the first challenge in robotics is to introduce a robot without friction into existing workflows,” Murty said. “So we have to be very careful about changing or remapping the workflows without destroying the profitability of the businesses.”
Maintaining profitability is a core tenet of the company’s strategy. Electric Sheep isn’t interested in waiting years for the company to start turning a profit, instead, it wants to maintain profits while also introducing robots that can reduce labor costs. Once Electric Sheep has revamped workflows to make them compatible with robots, they start rolling them out.
“[The next step] is to gradually introduce automation in the form of more crews,” Murty said. “For instance, what our transformation looks like is there’s a van with up to six to eight robots that get loaded in, and then they go around and do the mowing at different sites that were previously being manually mowed.”
Building autonomy for unstructured, outdoor environments
While the company is starting with robotic mowing, it has plans to automate more landscaping processes in the future.
“When we master mowing, turning on other service lines becomes a lot easier, because there are many other services that are also mechanized, meaning there’s a dude on top of a machine driving it around, and then there’s some type of accurate actuation,” Herold said.
Moving forward, the company plans to work on robots that can move snow, sweep parking lots, blow leaves, fertilize soil, and more. The company’s technology is built for outdoor unstructured environments, and its trying to build a foundational model similar to large language models like ChatGPT, but instead of language, the model would specialize in the physical world.
“Today, AI models really understand language, but they still have no sense of physical reality,” Herold said. “If you take a cat, it navigates the physical space effortlessly, you don’t have to train a cat to go around your house and do X, Y, Z. In the same way, I think the model that we’re building at its core is fundamentally semantically aware of its surroundings.”
Electric Sheep is working with a version of model-based reinforcement learning, and it’s building a Large World Model (LWM) that understands the structure of outdoor environments and how to act within them. After every day working in the world, the company’s robots upload real data, and Electric Sheep’s simulation engine runs different permutations of feedback loops at the scale of 1,000 robots. This helps the company’s robots rapidly improve with each day of work.
Electric Sheep’s business strategy and technology are closely tied together. At the core of the company is a belief that robotics companies need a new kind of business model to best utilize recent advances in machine learning.
“Collection and reinforcement learning has to happen at scale, which is why I think we’re sort of on this mission to prove how to scale that model as a very deliberate positioning and articulation,” Murty said. “This approach that we want to take is separate from, I’d say, 99.99% of robotics companies in the world who are trying to basically sell a robot to a customer.”