On May 13, Li Chang, founder of robotic lawn mower startup Oasa, published an open letter on social media. Despite pushing hard to deliver the company’s flagship R1 product, he wrote, Oasa had run out of steam. Team restructuring, manufacturing setbacks, and capital constraints ultimately brought operations to a halt.
Founded in August 2022, Oasa focused on developing smart lawn mowers and had previously raised nearly RMB 100 million (USD 14 million) in angel funding from Lanchi Ventures, Bear Electric, and other investors. Li was formerly a founding member of home robotics brand Narwal. In June 2024, Oasa launched its reel mower, the Oasa R1, starting at USD 1,199 on Kickstarter. The campaign raised around USD 2.3 million.
Oasa quickly funneled all its resources into product development and manufacturing. Yet there’s still no clear public update on how many crowdfunded units have actually been delivered. A person familiar with the matter told 36Kr that only a small batch had shipped.
For a robotics startup with heavy R&D investment, USD 2.3 million isn’t much. “Oasa pulled in USD 300,000 on the first day of crowdfunding, and that’s not a huge haul,” one industry insider told 36Kr. In an earlier interview, another founder in the sector put it bluntly: “If you want to build a lawn mower startup, you better have at least RMB 100 million ready.”
About two weeks before Li’s announcement, a customer who had paid a deposit for the R1 commented on Kickstarter that they had been unable to contact any Oasa developers for over a week following a software update on April 27. The company also hadn’t responded to Facebook messages. Days later, another backer said their order for a roller and blade attachment had been canceled and refunded, and asked for an explanation.
Rumors of Oasa’s collapse spread quickly, prompting widespread discussion among backers. The company remained silent, and its Amazon store links stopped working after May 10. Now, the news is confirmed: Oasa is shutting down.
Innovation meets limits
To its credit, the Oasa R1 introduced genuine product innovation.
It was the first lawn mower to use a reel blade system. Compared with traditional rotary blades, which often leave uneven cuts and flying grass clippings, reel blades roll and shear like scissors, creating cleaner, more uniform cuts with less damage to the turf.
The R1 also featured an automatic blade-sharpening system. This allowed users to skip manual inspections by monitoring blade wear in real time, reducing both the frequency and hassle of replacements or recalibrations.
But the drawbacks were just as evident. Reel blades are precision-engineered and expensive, with each set costing two to three times as much as a rotary system. The automatic sharpener adds even more complexity, requiring sensors, miniature actuators, and precise calibration. For a startup like Oasa in its early production stage, this led to immature supply chains and low yield rates, which drove up overall unit costs.
“You can’t just pitch a PowerPoint to build a product,” said Li Xian (pseudonym), a veteran mower engineer. “The reel blade concept was cool and the build looked solid, but just one set of blades already cost over RMB 1,000 (USD 140). It was too expensive.”
Technically, Oasa employed a hybrid solution combining vision and LiDAR (light detection and ranging). In lawn care robotics, performance is heavily shaped by the underlying algorithmic approach. The industry’s most common configurations include pure vision, vision with RTK (real-time kinematics), and vision with LiDAR:
- Vision-based systems are affordable and rich in contextual awareness but struggle in harsh lighting and with indistinct objects. Processing is also slower and less responsive.
- Vision plus RTK combines visual sensors with high-precision satellite positioning, offering strong reliability and boundary control. But it’s costly and depends on satellite and base station signals.
- Vision plus LiDAR brings accurate environmental mapping and performs well in varying light conditions. However, LiDAR sensors are expensive and have inherent blind spots.
According to one person familiar with Oasa’s product, the R1 used single-line LiDAR, a technique commonly found in robotic vacuums. “It’s actually very cheap, and during testing it didn’t work well under sunlight,” the person said.
“Sure, the blades were innovative, but the rest of the tech wasn’t quite up to speed. The differentiation wasn’t very meaningful,” added Xiao Yue (pseudonym), another lawn robotics engineer. For example, the R1 had two cameras and needed constant data training to improve, which meant long lead times for machine learning optimization.
Founder Li understood the importance of foundational tech. At its peak, Oasa employed nearly 50 people, and software engineers reportedly drew annual salaries of RMB 700,000–800,000 (USD 98,000–112,000). These high personnel costs only deepened the company’s financial strain.
In his letter, Li acknowledged that mounting operational and financial pressures had overwhelmed the startup’s ability to stay afloat.
Ultimately, while Oasa delivered structural innovation, lawn robots are deeply integrated systems, and any weak link in the hardware-software matrix can collapse the entire product. Without a full stack of engineering strength, even standout features become liabilities.
Choke points
Globally, the lawn care robot market is heavily concentrated in the US and Europe, where sprawling backyards drive demand.
Statista estimates there are about 250 million gardens worldwide. The US leads with around 100 million, or 40% of the global market, followed by Europe with roughly 80 million. While robot mower penetration is 10–15% in Europe, it’s just 2% in the US, where many still use handheld mowers.
According to customs data, Germany tops China’s export list for lawn mowers, followed by the UK and then the US.
“Europe sees higher volume. Many users are switching from wired mowers to wireless models,” said Peng Li (pseudonym), head of a mower startup.
Oasa’s R1 primarily targeted Germany and the US. But Germany is fiercely competitive, dominated by legacy players like Husqvarna and Positec. Positec alone holds 59% market share in the country. New entrants like Ninebot and Ecovacs are also rapidly gaining ground.
“Ninebot already had brand equity, so distributors in Europe and North America were reaching out to them proactively to run product tests,” Peng Li said. “For companies like mine, most of our channel partners are still Chinese.”
Whoever secures the best local partners first gains a real edge. According to Li Xian, Oasa lacked strong channel capabilities in Germany, limiting its market reach from the start.
Meanwhile, Oasa’s US ambitions were also constrained by scale. About 40% of US homes have lawns larger than 1,000 square meters, with many featuring open, uneven terrain such as wet grass and woodlands. Typical middle-class homes sit on plots between 1,300–2,000 square meters.
The R1 could only handle lawns up to 500 square meters, with a theoretical max of 1,000. In contrast, Agilex’s Luba 2 AWD and Yuka models support areas over 3,000 square meters. The Lymow One can cut 7,000 square meters per day. Yarbo M1 from Hanyang Robotics maxes out at six acres.
So in markets like the US and Europe, where big backyards are the norm, Oasa’s limited range was a clear disadvantage.
Over the past two years, lawn robotics has surged in popularity alongside pool cleaners and snow-removal bots. But the boom has also raised the stakes. Many startups are feeling the squeeze.
This isn’t the first wave of casualties. As early as 2023, the field saw companies collapse and projects shutter. Anker Innovations disbanded its mower team after failing to compete. EcoFlow’s mower didn’t meet expectations and was eventually pulled. Heisenberg Robotics went into default and was listed as a credit risk.
Oasa’s abrupt fall may be a reality check for the sector, but it’s also a sign that to review the balance between technical ambition and commercial viability. The industry may now be entering a more sober phase.
This article was written by Huang Nan and was originally published by 36Kr.
