EoT · Project

Always-On Embedded Computer Vision Platform for Smart Devices and Surveillance

digitalTestedTRL 6

Imagine a tiny, ultra-low-power camera brain that can "see" and understand what's happening around it — without draining a battery in minutes. Your phone camera is powerful but dies fast if you leave it running all day. This team built a miniature vision board that stays on continuously, fits inside everyday objects, and makes sense of what it sees in real time. Think of it as giving cheap, tireless "eyes" to everything from security cameras to wearable gadgets to factory robots.

By the numbers

EUR 3,734,830

EU funding for platform development

consortium partners

countries in consortium

SMEs in the consortium

78%

industry partner ratio

total project deliverables

The business problem

What needed solving

Most computer vision systems today face a hard trade-off: they're either portable but drain batteries fast, or they run continuously but need fixed power and bulky hardware. Businesses wanting to add "smart eyes" to products, buildings, or wearables hit a wall — existing mobile devices waste power on unused sensors and weren't designed for always-on vision. This leaves entire categories of applications (continuous monitoring, wearable AR, embedded inspection) without a viable low-power solution.

The solution

What was built

The project built a physical embedded vision device with a dedicated form factor board — a compact, low-power hardware platform purpose-built for continuous computer vision. They also delivered open APIs designed to maximize visual information extracted per milliwatt, plus an application development competition framework to build an ecosystem around the platform. In total, 27 deliverables were produced across the project.

Audience

Who needs this

Video surveillance and smart security companies needing always-on, low-power camera intelligenceWearable device manufacturers adding visual capabilities to glasses, helmets, or body-worn camerasIndustrial inspection firms building portable or drone-mounted visual quality control toolsSmart building integrators embedding vision into lighting, HVAC, or access control systemsAssisted living technology companies developing visual monitoring for elderly care

Business applications

Who can put this to work

Security & Surveillance

SME

Target: Video surveillance system integrators and smart building companies

If you are a surveillance company struggling with high energy costs and limited intelligence in your camera networks — this project developed an always-on embedded vision device that processes video locally without cloud dependency. With 9 consortium partners across 7 countries and a physical prototype board delivered, the platform could replace power-hungry camera systems with compact, energy-efficient units that run computer vision algorithms continuously.

Retail & Smart Spaces

mid-size

Target: Retail analytics and customer experience technology providers

If you are a retail tech company looking to add visual intelligence to stores without expensive cloud infrastructure — this project built a low-power vision platform that can be embedded into shelves, signage, or kiosks. The device infers information from visual data at minimal power consumption, enabling people counting, heat mapping, or product interaction tracking without always-connected internet. The consortium included 6 SMEs with direct industry experience.

Industrial Automation & Robotics

any

Target: Machine vision and quality inspection system manufacturers

If you are an industrial automation company needing vision capabilities in mobile robots or portable inspection tools — this project created an open hardware vision platform optimized for maximum inferred information per milliwatt. With 27 deliverables produced and a dedicated device board, it addresses the gap where current systems are either mobile or always-on, but not both. The EUR 3,734,830 EU-funded effort specifically targeted embedded vision beyond factory automation.

Frequently asked

Quick answers

What would it cost to license or integrate this vision platform into our products?

The project was funded as an Innovation Action with EUR 3,734,830 in EU contribution across 9 partners. Licensing terms would need to be negotiated directly with the consortium, led by Universidad de Castilla - La Mancha. As an open hardware platform, some components may be available under open-source terms — contact the coordinator for specifics.

Can this scale to thousands of units in a production environment?

The project delivered a physical device with a form factor board, indicating hardware moved beyond lab prototype. With 6 SMEs and 78% industry ratio in the consortium, the design was oriented toward commercial viability. However, mass production readiness would need to be verified with the current state of the hardware design post-project.

What is the IP situation — can we use this technology freely?

The consortium of 9 partners across 7 countries jointly developed the platform. IP ownership and licensing rights are governed by the Horizon 2020 grant agreement. The project emphasized open hardware and APIs, but specific patent or licensing terms should be clarified with the coordinator at Universidad de Castilla - La Mancha.

How does this compare to existing edge AI solutions like NVIDIA Jetson or Google Coral?

EoT was specifically designed to maximize inferred information per milliwatt — prioritizing ultra-low power over raw performance. Unlike general-purpose edge AI boards, this platform was built from the ground up for continuous vision applications. Based on available project data, it targets use cases where always-on operation matters more than peak processing speed.

Is this technology ready to deploy today or still experimental?

The project closed in June 2018 and delivered a working device board plus 27 total deliverables. As an Innovation Action (not basic research), it targeted near-market readiness. However, given the project ended over 7 years ago, current deployment readiness depends on whether consortium partners continued development commercially.

What specific vision tasks can this platform perform?

Based on the project objective, the platform supports computer vision applications including wearable augmented reality, surveillance, and ambient-assisted living. The APIs were designed to adapt quality of inferred results to each application's needs. Specific algorithm capabilities would need to be confirmed with the consortium.

Consortium

Who built it

The EoT consortium is heavily industry-driven with 7 out of 9 partners from industry and 6 being SMEs — giving it a 78% industry ratio, which is unusually high for EU research projects. This signals the technology was built with commercial intent, not just academic curiosity. The single university partner (Universidad de Castilla - La Mancha, Spain) served as coordinator, providing the core research capability, while 1 research organization rounded out the team. Geographic spread across 7 countries (Austria, Switzerland, Germany, Spain, France, Ireland, Portugal) gives the consortium broad European market access. For a business considering this technology, the high SME count means there are likely multiple smaller companies already exploring commercialization paths for components of this platform.

UNIVERSIDAD DE CASTILLA - LA MANCHACoordinator · ES
NVISO SAparticipant · CH
THALES SIX GTS FRANCE SASparticipant · FR
EVERCAM LIMITEDparticipant · IE
DEUTSCHES FORSCHUNGSZENTRUM FUR KUNSTLICHE INTELLIGENZ GMBHparticipant · DE
FLUXGUIDE AUSSTELLUNGSSYSTEME GMBHparticipant · AT

How to reach the team

Universidad de Castilla - La Mancha (Spain) — reach out to the Computer Architecture department or the project's principal investigator via the university directory

Order a report on this consortium See UNIVERSIDAD DE CASTILLA - LA MANCHA profile →

Next steps

Talk to the team behind this work.

Want an introduction to the EoT team? SciTransfer can connect you with the right person in the consortium who matches your specific use case — whether that's the hardware designers, the API developers, or the SME partners already working on commercial applications.

Order a report on this topic More in Digital & ICT