Coordinated APRICOT (€2.48M) to develop an Artificial Perception Chip for the Intelligence of Things.
ANOTHER BRAIN
Paris-based AI deep-tech SME building perception chips and embedded machine learning for IoT devices and automotive thermal vision.
Their core work
ANOTHER BRAIN is a Paris-based deep-tech SME building artificial perception systems — chips and software that let machines see, interpret, and act on visual information without cloud dependency. Their H2020 work centers on two complementary fronts: an AI perception chip for Internet-of-Things devices (APRICOT) and thermal-vision machine learning for autonomous driving (HELIAUS). In practical terms, they turn raw sensor input (standard cameras, thermal imagers) into real-time decisions on embedded hardware, serving automotive, IoT, and industrial customers that need on-device intelligence. They are a technology provider, not a research lab — the output is silicon, firmware, and ML models ready for integration.
What they specialise in
Both APRICOT and HELIAUS target on-device ML in resource-constrained hardware rather than cloud inference.
Partnered in HELIAUS on thermal vision augmented awareness for autonomous vehicles.
HELIAUS applied machine learning to thermal imagery for autonomous driving applications.
APRICOT's chip focus and HELIAUS's embedded-systems keyword indicate a consistent edge-compute orientation.
How they've shifted over time
With only two H2020 projects one year apart, the evolution is a narrowing rather than a pivot. In 2018 they launched as coordinator of APRICOT, defining a broad IoT perception-chip agenda. By 2019 they had attached that core capability to a concrete automotive application — thermal vision for autonomous driving in HELIAUS — shifting from generic IoT to a named vertical with a paying end-market.
They are moving from horizontal chip development toward vertical integration in automotive perception, making them a plausible partner for anyone building ADAS, driver-monitoring, or thermal sensing products.
How they like to work
They have shown they can lead (coordinating the larger APRICOT grant under the SME-2 instrument) and also slot in as a specialist partner in a research-action consortium (HELIAUS). With 11 unique partners across just two projects, each engagement brings a fresh network rather than a repeated inner circle — suggesting an opportunistic, capability-led collaboration model rather than a closed club.
Eleven distinct consortium partners spread across four countries, with the anchor firmly in France (Paris HQ). Their network footprint is modest in breadth but concentrated on Western European industrial and research players typical of automotive and semiconductor consortia.
What sets them apart
Few French SMEs combine their own perception chip development (APRICOT) with applied machine-learning deployment on thermal sensors (HELIAUS) — most players do one or the other. Partnering with them means accessing a single team that can take a vision problem from silicon architecture through ML model to embedded deployment. That vertical stack is what distinguishes them from pure-play chip designers and from software-only AI startups.
Highlights from their portfolio
- APRICOTTheir largest grant (€2.48M) and only coordinator role — a clear signal that IoT perception chips are their flagship capability.
- HELIAUSMoves their perception technology out of the lab into autonomous driving with a multi-country industrial consortium.