Home security cameras are becoming part of a larger debate about privacy, cost and the role of remote servers in everyday devices. Plumerai, a London-based company established in 2017, is trying to shift more of that intelligence onto the camera itself.
The company’s work centers on tiny AI: smaller models designed to run on low-cost, low-power chips rather than relying on large remote systems. Its latest step is a deal with Chamberlain Group, the corporate parent of brands including myQ and LiftMaster, which will use Plumerai’s technology in smart cameras beginning with an outdoor cam.
Why on-device AI matters for cameras
The privacy pressure around cameras has become harder to ignore. The source article points to Harvard students who made headlines on Wednesday by adding facial recognition to Ray-Ban Meta glasses, a project that added to existing concerns about widely available camera technology.
Security cameras raise a related issue: where the video and analysis happen. If a device needs to send data to a remote server to understand what it sees, that adds another layer of security and privacy concern. The article also notes that more issues arise when Ring’s parent Amazon and law enforcement enter the picture.
Plumerai’s core pitch is that common camera intelligence can run locally. The company says its technology can handle tasks such as people detection and familiar face identification without sending data to a remote server.
That distinction matters because the camera is no longer just a lens connected to the internet. It is also a computing device. If more of the analysis happens on the hardware in the home, the system depends less on cloud processing for basic AI features.
The cost problem behind cloud processing
Tony Fadell, an early investor in Plumerai and the creator of the iPod, connects the problem to his experience as co-founder of Nest. He told TechCrunch, “We’d have to worry so much just about the storage cost and the data transmission costs,” adding, “We’re taking full frames. It’s a ton of stuff that we’re recording, but not recording on-camera. I felt the weight of this all the time.”
That cost pressure is not only a company problem. When products require additional computers, storage and transmission, the added expense is often passed on to consumers. Fadell points to Ring’s recent decision to double its professional 24/7 monitoring costs as a key indicator.
On-device AI offers a different path. If the camera can identify people or familiar faces locally, it can reduce the need for constant remote processing. The source does not claim this removes every cost or privacy issue, but it frames local intelligence as a way to address some of the pressure created by cloud-heavy systems.
How tiny AI differs from large models
Plumerai specializes in tiny AI rather than the large language models associated with platforms like ChatGPT. The article describes those large models as built on vast stores of data, requiring too much computing power for a small consumer electronics device and being prone to hallucinations.
For a home security camera, the job is more focused. The device does not need to generate long answers or handle a wide range of language tasks. It needs to recognize relevant visual events reliably while running on hardware with limited power and cost constraints.
Fadell compares the smaller-model strategy to the development path from the iPod to the iPhone. “The only reason the iPhone could exist is because we started small with the iPod. Usually you can grow things up, you can’t make big things small,” he says. “So we started really small and grew the iPhone up f rom that. Remember, Micr osoft tried to take Windows and make Windows Mobile on a phone. They take this big thing and it never worked. You have to start small, and then you grow from there.”
The point is not that every AI system should be small. It is that some products need AI designed around the constraints of the device from the beginning. For smart home cameras, Plumerai is arguing that smaller, specialized models are the better fit.
Chamberlain brings Plumerai into smart cameras
Plumerai CEO Roeland Nusselder says the company has been building this technology for a long time. “If you look at it empirically, our tiny AI is more accurate and runs on lower cost, lower power chips than anything else that’s out there — especially in the smart home camera market.”
The Chamberlain Group partnership gives Plumerai a concrete route into consumer hardware. Chamberlain, based in Illinois, will incorporate Plumerai’s offering into its smart cameras, starting with an outdoor cam.
Nusselder describes the integration directly: “All of the AI features are from Plumerai, running locally on the camera,” he says. “How I look at Chamberlain is a company that’s not a Big Tech company, but that’s able to achieve really great things with small AI.”
That positioning is central to the story. Plumerai is not presented as a large platform company trying to spread one giant AI system across many product categories. It is a focused startup working in a specific segment of the smart home camera market.
Small teams and focused AI
The article notes that Plumerai has not disclosed its headcount, but says it is almost certainly well below the teams behind Ring and Nest. That smaller scale is part of the company’s identity: it has focused on one market segment rather than trying to cover the broad range of products owned by massive corporations such as Amazon and Google.
Fadell, who has worked as an executive at some of the world’s largest tech firms, now spends time helping startups including Plumerai. His view is that focused teams can move quickly when they have the right expertise.
“Focus is the key,” he explains. “I have learned that small teams — in the 10s, the 50s — can really do a lot when it’s the right set of expertise around the table. I like to be at the tip of the spear on disruptive technologies. It’s small teams with the right idea.”
For home security cameras, the stakes are practical. People want useful features such as people detection and familiar face identification, but the way those features are delivered affects privacy, infrastructure and cost. Plumerai’s bet is that tiny AI running on the device can make smart cameras more capable without making them more dependent on remote servers.