Executive summary — the one-minute version
|
🏭 A design that reached mass production Sticker-Eye went from design win to shipped hardware, launched through a public campaign that closed at about 18× its funding goal — evidence that the design, BOM and supply chain hold up at volume. |
🧠 AI runs on the camera itself An optimized embedded neural network performs person detection and face recognition on the device, rejects non-human motion and keeps working with no internet connection. |
|
🔋 Six months on one charge Low-power system engineering — a 6,000 mAh Li-ion pack, energy-saving AI logic and a tight power budget — delivers up to six months of standby between USB Type-C charges. |
🔌 Open where it matters The device side carries the interfaces for cloud connection and for the iOS/Android app. Detection and recording run locally; the cloud and app layers are integration points the customer brands and operates. |
A cordless AI camera looks simple on a spec sheet and is hard to build. Every function — sensing, detection, storage, connectivity — has to fit inside a battery budget, because there is no mains power to fall back on. Detection has to run at the edge: a camera that depends on a cloud round-trip burns its battery on radio traffic and goes blind when the connection drops. And the mechanics have to turn installation into a seconds-long, no-tools task on any surface, indoors or out.
EyeCloud took on the full problem as one program: sensing and optics, embedded AI, low-power hardware, mechanical design, sealing, and manufacturing. The result is Sticker-Eye — a shipped product that doubles as the reference for what our smart camera overall solution can do for your product.
| Imaging | 1080p video day and night, 110° viewing angle, night vision up to 16 ft (~5 m), person-triggered video recording |
| On-device AI | Optimized embedded neural network: person detection and face recognition executed locally; energy-saving inference scheduling to protect the battery budget |
| Power | Built-in 6,000 mAh Li-ion rechargeable pack, USB Type-C charging, up to six months of standby depending on settings, usage and temperature |
| Mechanics | Sticky pad + magnetic stand (three optional screws); sets on any horizontal surface or affixes to walls and ceilings; upside-down mounting supported by a re-orienting camera head |
| Enclosure | Weather-sealed, weatherproof construction for outdoor use |
| Privacy | Encrypted local storage; on-device processing that does not require an internet connection |
| Connectivity | Device-side interfaces for cloud connection and for the iOS/Android app (see Built to connect) |
The problem: in battery-powered cameras, most false triggers come from falling leaves, animals, cars and other motion inside the field of view — and every false trigger costs battery. Cloud-loop detection makes both problems worse: it burns radio time and fails when the network does.
What EyeCloud engineered: an optimized embedded neural network that sees and analyzes objects locally, triggers recording for people, recognizes faces and skips everything else. Running detection on the device is what keeps the camera useful with no internet at all, and it is the main lever on battery life. Getting this to fit inside a sealed, battery-powered enclosure is exactly the kind of embedded AI engineering EyeCloud’s team — with its Intel and Movidius background in computer vision, image sensors and AI algorithms — applies on customer programs.
No drills, no wall anchors, no cable runs: a double-sided sticky circle, a magnetic stand and three optional screws. The body affixes to a vertical wall, mounts upside down under a ceiling or eave, stands on a shelf, or clamps to a pole or tree — and the camera head re-orients so the picture stays level. For a brand, this is a serviceability and returns story as much as an installation one: end users can reposition the camera themselves, in seconds, without breaking anything.
Sticker-Eye processes what it sees on the device and stores footage encrypted locally; detection works without an internet connection. Video never has to leave the device — which is both a privacy posture and a compliance simplification for whichever brand operates the service around it.
Before entering volume production, the platform was validated publicly: a crowdfunding campaign closed at $181,405 from 1,016 backers — roughly 18× its $10,000 goal — and hardware shipped to customers worldwide. For our team it doubled as a full-cycle rehearsal of everything a camera program depends on beyond the schematic: positioning, packaging, certification and worldwide logistics, executed on a real product at real volume.
Sticker-Eye is deliberately device-first: detection, recording and storage happen on the camera, and the camera is fully useful on its own. For everything beyond that, the device side ships with the two integration interfaces a smart-camera project almost always needs:
Those interfaces are plumbing, not the product — and that is exactly the point for our customers. A brand can bring its own cloud and its own app, plug the Sticker-Eye device in through these interfaces, and ship it under its own name; EyeCloud’s scope of responsibility is the device and the overall solution around it.
Sticker-Eye is a reference, not a fixed product. The blocks that produced it are the capability set we apply to customer programs, adapted to your use case, brand and bill of materials:
| Sensing & optics | Image sensor selection, lens matching and module design, with image quality tuning for the target scenes |
| Embedded AI | Person and face detection models optimized for low-power SoCs, plus energy-saving inference scheduling |
| Low-power hardware | Battery pack and charging design, power budgeting across sensing, inference and connectivity |
| Mechanics & industrial design | Mounting mechanics, weather sealing and enclosure design adapted to your form factor and brand identity |
| Integration | Device-side cloud and app interfaces wired to your cloud platform and mobile app |
| Production | R&D and smart manufacturing in Hangzhou, supply chain and production support from Korea, shipment worldwide — the same team backing the other EyeCloud programs |
It is an EyeCloud overall solution for a low-power AI camera: industrial design, embedded AI hardware, weatherproof mechanics, manufacturing, and the device-side interfaces for cloud and app connectivity — delivered as one program and proven by a shipped, market-validated product.
Yes. Scope, industrial design, AI tuning and integration are adapted per program; Sticker-Eye’s platform blocks — enclosure and mounting mechanics, low-power AI stack, integration interfaces, manufacturing — are the starting point, not the limit.
An optimized embedded neural network analyzes objects locally and triggers on people, so falling leaves, animals, cars and other moving objects do not fire alerts or recordings. Face recognition also runs on the device.
No. Detection and recording run entirely on the camera, and footage is stored encrypted locally. The internet (and the cloud/app interfaces) only comes into play for remote access and livestream.
A 6,000 mAh rechargeable Li-ion pack charged over USB Type-C, combined with energy-saving AI logic and a power budget tuned across sensing, inference and connectivity. Depending on settings, usage and temperature, the camera reaches up to six months in standby mode.
Yes — that is the intended path. The device exposes the interfaces for cloud connection and for an iOS/Android app, and EyeCloud’s job is to wire the device side to your stack.
Scope, design, AI tuning, integration, pilot and mass production, run by one team: R&D and smart manufacturing in Hangzhou, supply chain and production support from Korea, and delivery worldwide.
Sticker-Eye shows what EyeCloud delivers as an overall solution: concept, embedded AI, weatherproof design, manufacturing, and open device-side interfaces for your cloud and app. Explore our cameras and edge AI products, browse our other engineering case studies — such as the AI-SOP assembly verification case — or contact our team to scope your program.
