Automatika Robotics
Software infrastructure for physical AI agents
Website: https://automatikarobotics.com/
Cover Block
PUBLIC
| Attribute | Detail |
|---|---|
| Name | Automatika Robotics |
| Tagline | Software infrastructure for physical AI agents |
| Headquarters | France |
| Founded | 2022 |
| Stage | Pre-Seed |
| Business Model | B2B |
| Industry | Deeptech |
| Technology | Robotics |
| Geography | Western Europe |
| Founding Team | Academic Spinout (Inria) |
| Funding Label | Undisclosed |
Links
PUBLIC
- Website: https://automatikarobotics.com/
- LinkedIn: https://fr.linkedin.com/company/automatika-robotics
- GitHub: https://github.com/automatika-robotics
Executive Summary
PUBLIC Automatika Robotics is building the software infrastructure layer for general-purpose physical AI agents, a bet that hinges on enabling robots to understand and navigate complex, human-populated environments [Automatika Robotics, 2026]. The company is an academic spinout from Inria, France's national research institute for computer science, founded in early 2022 by researchers specializing in robotics and machine learning [Freeway Ecosystem]. Its primary offering, the EMOS platform, is designed to provide a unified environment for engineering physical AI, with its open-source Kompass navigation stack representing a tangible, documented component of this broader architecture [Automatika Robotics, 2026].
The founding team's background is rooted in deep tech research, with co-founder Haroon Rasheed's prior work at Inria's Morpheo team providing a direct link to advanced computer vision and 3D modeling research [Inria, 2026]. While the company has participated in the Techstars Torino Cities of the Future accelerator, its capitalization remains undisclosed, with no public funding rounds, named customers, or commercial traction signals available in primary sources [Techstars, 2026]. The business model is presumed B2B, targeting robotics hardware makers and deployers, but revenue mechanics are not detailed.
Over the next 12-18 months, the critical watchpoints are the transition from open-source components to a commercialized EMOS platform, any disclosed seed funding to validate investor interest, and the emergence of initial deployment partners to move beyond pure research and development. Data Accuracy: YELLOW -- Company claims and academic background are documented; commercial and financial status is unverified.
Taxonomy Snapshot
| Axis | Classification |
|---|---|
| Stage | Pre-Seed |
| Business Model | B2B |
| Industry / Vertical | Deeptech |
| Technology Type | Robotics |
| Geography | Western Europe |
| Founding Team | Academic Spinout |
Company Overview
PUBLIC
Automatika Robotics was founded in 2022 as a spinout from Inria, France's national research institute for computer science and applied mathematics [Freeway Ecosystem]. The company is headquartered in France, though its specific legal entity structure is not detailed in public sources. The founding story centers on a team of researchers specializing in robotics, computer vision, control theory, and machine learning aiming to translate academic work into commercial software infrastructure for physical AI agents [Automatika Robotics, 2026].
A key early milestone was the company's participation in the 2023 cohort of the Techstars Torino Cities of the Future Accelerator, a program focused on smart mobility and urban innovation [Techstars, 2026]. This affiliation is listed in the Techstars jobs directory and was announced in the accelerator's cohort unveiling. Since its founding, the primary public-facing development has been the release of its open-source software components, including the Kompass navigation stack documentation and the EmbodiedAgents framework on GitHub, which hosts 12 repositories [GitHub, 2026].
No subsequent funding announcements, customer deployments, or major press coverage have been identified since the 2023 accelerator participation. The company's website and documentation were last updated with a 2026 copyright notice, but no news or blog posts detailing commercial progress are present.
Data Accuracy: YELLOW -- Founding year and Techstars participation are confirmed; academic spinout claim is partially corroborated via founder background. No independent verification of legal entity or detailed milestone timeline.
Product and Technology
MIXED The company's public technical footprint is anchored by two distinct but related software components: a commercial platform and an open-source navigation stack. The EMOS platform is described as the primary offering, a single platform "to engineer physical AI" [Automatika Robotics website, 2026]. The company's public-facing documentation directs users to EMOS for usage details, positioning it as the core commercial infrastructure product [Automatika Robotics, 2026].
Alongside EMOS, Automatika maintains Kompass, a navigation stack with publicly available developer documentation for version 0.4.1 [Automatika Robotics, 2026]. The Kompass documentation details a component-based architecture for robotics, including controllers, actions, and launchers, suggesting a framework built for extensibility by developers [Automatika Robotics, 2026]. A key open-source component is the EmbodiedAgents framework, hosted on the company's GitHub, which is described as a fully-loaded ROS2 framework enabling interactive physical agents to understand, remember, and act upon contextual information [GitHub, 2026]. This repository, alongside 11 others under the Automatika Robotics organization, represents the most tangible and active output from the team [GitHub, 2026].
The technical narrative emphasizes building general-purpose intelligence for robots, with a stated focus on enabling comprehension of space, human behavior, and safe collaboration [LinkedIn]. The initial technical wedge appears to be advanced navigation systems, targeting applications in autonomous driving and embodied AI [Freeway Ecosystem]. No public demos, case studies, or detailed technical whitepapers beyond the API documentation and GitHub repositories were identified.
Data Accuracy: YELLOW -- Product claims are from the company's own website and GitHub; Kompass documentation is publicly accessible. No independent verification of EMOS platform functionality or deployment.
Market Research
MIXED
The ambition to build general-purpose physical AI agents is moving from academic labs into venture-backed startups, but the commercial infrastructure layer for these systems remains nascent. Automatika Robotics targets this specific wedge: the software platform that sits between foundational AI models and the physical hardware of robots, enabling them to perceive, navigate, and act in unstructured environments.
Quantifying the total addressable market for a pre-product robotics software stack is speculative. No third-party reports specifically sizing the "physical AI agent infrastructure" market were identified in the research. Analysts can draw an analogous market from the broader commercial robotics software sector. According to a 2023 report from MarketsandMarkets, the global robot software market was valued at $13.8 billion and is projected to reach $43.2 billion by 2028, growing at a CAGR of 25.6% [MarketsandMarkets, 2023]. While this encompasses a wide range of software from simulation to fleet management, it provides a relevant upper-bound context for Automatika's category.
Demand for such infrastructure is driven by several converging trends. The maturation of large language and multimodal models has created a new cognitive layer for robots, but integrating these models with real-time sensor data, safety-critical control systems, and legacy robotic middleware like ROS2 presents a significant engineering challenge. This creates a clear wedge for a dedicated software layer. Furthermore, labor shortages in sectors like logistics, manufacturing, and healthcare continue to push investment into automation, with a growing focus on more flexible, AI-driven systems rather than single-task industrial arms.
Key adjacent and substitute markets include traditional industrial automation software, autonomous vehicle stacks, and the broader AI developer tools ecosystem. The risk for a focused infrastructure play is being subsumed by larger platforms; for instance, NVIDIA's Isaac platform already offers a comprehensive suite for robot simulation and development. However, the technical specificity required for safe human-robot collaboration in dynamic settings suggests there may be room for specialized, best-of-breed solutions.
Regulatory and macro forces are a double-edged sword. In Europe, the proposed AI Act and machinery regulations will impose strict requirements on safety and transparency for autonomous systems, potentially raising the compliance bar for new entrants. Conversely, these regulations could also act as a moat for companies that build certified, safety-by-design frameworks from the outset, a potential advantage for an academic spinout with a research pedigree.
Robot Software Market 2023 | 13.8 | $B
Robot Software Market 2028 (projected) | 43.2 | $B
The projected growth of the broader robot software market underscores the underlying tailwind, but it does not confirm demand for Automatika's specific architectural approach. The company's success will depend on its ability to carve out a definable serviceable obtainable market (SOM) within this expansive landscape, likely starting with research institutions and early-adopter robotics OEMs.
Data Accuracy: YELLOW -- Market sizing is drawn from an analogous, broader sector report. Company-specific TAM/SAM/SOM is not publicly available.
Competitive Landscape
MIXED
Automatika Robotics is a technical infrastructure play entering a market defined by large, well-funded incumbents in robotics middleware and a fragmented landscape of specialized AI model providers.
Given the absence of named competitors in the captured sources, a direct comparison table cannot be constructed. The competitive analysis must be drawn from the broader market context surrounding the company's stated focus on software infrastructure for physical AI agents.
The competitive map for robotics software infrastructure is stratified. At the platform layer, incumbent open-source frameworks like ROS (Robot Operating System) and its successor ROS 2 represent the de facto standard, backed by large commercial entities like Open Robotics and supported by a vast ecosystem. These are not direct competitors but the foundational layer upon which Automatika's EMOS and Kompass would need to interoperate or replace. The challenger segment includes well-funded startups such as Formant (raised $18M in 2021) and Foxglove (raised $12M in 2022), which focus on robot data management and visualization, positioning themselves as the 'UI layer' rather than the core intelligence stack. In the adjacent substitute category, large cloud providers (AWS RoboMaker, Google Cloud Robotics Core) and industrial automation giants (Siemens, Rockwell Automation) offer integrated suites that could absorb the need for a standalone navigation and simulation infrastructure.
Where Automatika might claim a defensible edge today is in its academic pedigree and early-stage technical focus. The spinout from Inria provides access to deep research in computer vision and control theory, a talent pipeline that is difficult for generalist software firms to replicate quickly. This edge is currently perishable, however, as it is not yet commercialized or protected by significant intellectual property filings or a proprietary dataset. The company's early open-source activity, with 12 GitHub repositories including the EmbodiedAgents framework, suggests a strategy of building developer mindshare, a classic durable advantage in infrastructure markets if it gains traction.
The company's most significant exposure is to capital and commercialization speed. It operates with no disclosed funding in a sector where credible competitors have raised tens of millions. A named challenger like Formant has already established enterprise customers and a clear revenue model, advantages Automatika cannot currently match. Furthermore, the company does not own a distribution channel or a hardware partnership that would guarantee deployment of its stack, leaving it vulnerable to being bypassed by integrated solutions from larger players.
The most plausible 18-month scenario hinges on the company's ability to transition from research to a commercial proof point. The winner in this niche will likely be the first to demonstrate a scalable, general-purpose navigation stack adopted by a major robotics OEM or deployed in a high-profile autonomous vehicle pilot. If Automatika can use its Techstars network to secure a seed round and a flagship partnership, it could emerge as a specialist contender. The loser, conversely, will be any team that remains in perpetual stealth or research mode, as the market is moving rapidly with increased investment in embodied AI. Without a clear commercial milestone within this timeframe, Automatika risks being relegated to an interesting research artifact rather than a commercial entity.
Data Accuracy: YELLOW -- Competitive context is inferred from the broader market; no direct competitors are named in company sources.
Opportunity
PUBLIC The prize for Automatika Robotics is the foundational software layer for a new generation of interactive, context-aware physical agents, a role that could command platform-level economics if the company can transition from a research project to a commercial standard.
The headline opportunity is to become the default infrastructure provider for embodied AI, analogous to what ROS (Robot Operating System) became for traditional robotics research but with a native focus on higher-level intelligence and human interaction. The company's core mission, as stated on its website, is "to create the software infrastructure that empowers intelligent physical agents to operate seamlessly in the real world" [Automatika Robotics, 2026]. This outcome is reachable not because of current traction, but because of the technical foundation being laid in public. The company has published a fully-loaded ROS2 framework called EmbodiedAgents, described as enabling agents that "understand, remember, and act upon contextual information" [GitHub, 2026]. This open-source component, alongside the Kompass navigation stack documentation, represents a tangible, technical starting point for a platform, moving beyond conceptual marketing into developer-facing tools.
Growth Scenarios
Three concrete paths exist for Automatika to scale from an academic spinout to a significant commercial entity.
| Scenario | What happens | Catalyst | Why it's plausible |
|---|---|---|---|
| The Research-to-Industry Bridge | Automatika's EMOS platform becomes the preferred simulation and testing environment for robotics labs and hardware startups, monetized via enterprise licenses. | A major partnership with a European research consortium or a hardware OEM (e.g., Boston Dynamics, ANYbotics) to co-develop and standardize. | The team's roots in Inria, a premier European research institute, provide a natural network into academic and industrial research partnerships [Freeway Ecosystem]. The focus on "simulation, perception, and navigation" aligns with critical pre-deployment needs [Automatika Robotics]. |
| The Navigation Module Wedge | Kompass evolves into the dominant, standalone navigation stack for autonomous mobile robots (AMRs) in logistics and manufacturing, sold as a high-margin software SDK. | Securing a design-win with a single high-volume AMR manufacturer, validating performance in a real warehouse deployment. | The company explicitly cites an "initial wedge is advanced navigation systems for autonomous driving and embodied AI" [Freeway Ecosystem]. Navigation is a discrete, critical problem with a clearer path to integration and ROI than a full-stack "agent" platform. |
| The Agent Framework Standard | EmbodiedAgents becomes the de facto open-source framework for building interactive physical AI, with Automatika monetizing through cloud-hosted services, managed deployments, and premium support. | Achieving critical mass of developer adoption on GitHub, followed by the launch of a commercial cloud offering (EMOS Cloud). | The EmbodiedAgents repository is publicly available and positioned as a "core EMOS component" [GitHub, 2026]. Successful open-source models in adjacent spaces (e.g., OpenAI's GPT ecosystem, Hugging Face's transformers) demonstrate the viability of this adoption-led path. |
What compounding looks like hinges on a classic open-source playbook evolving into a data and distribution moat. Early adoption of the open-source frameworks (Kompass, EmbodiedAgents) by developers and researchers creates a feedback loop. User contributions and real-world testing improve the software, making it more robust and attracting more users. This growing community then creates a distribution channel for Automatika's commercial products, whether that's EMOS licenses or cloud services. The data moat forms if the company's platforms begin to aggregate unique datasets from diverse physical deployments,datasets on navigation failures, human-robot interaction patterns, or simulation-to-real-world gaps,that can be used to train superior proprietary models. The evidence of this flywheel starting is minimal, but the existence of 12 GitHub repositories suggests an active, if early, development posture aimed at building that initial community [GitHub, 2026].
The size of the win can be framed by looking at a credible comparable: the valuation of companies building foundational software layers for adjacent compute paradigms. While no direct public peer exists for embodied AI infrastructure, UiPath, which automates software processes, reached a market cap exceeding $10 billion post-IPO. A more specialized but relevant comparison is Nvidia's Isaac robotics platform, whose strategic value is embedded in its drive to sell GPU hardware. If Automatika successfully executes on the "Agent Framework Standard" scenario and captures a meaningful portion of the emerging embodied AI developer ecosystem, an outcome in the hundreds of millions to low single-digit billions of dollars is plausible (scenario, not a forecast). This is supported by the broader market context where investors are allocating significant capital to "AI that acts in the physical world," as seen in rounds for companies like Covariant and Figure AI, though Automatika's focus is squarely on the software layer beneath such applications.
Data Accuracy: YELLOW -- Scenario analysis is speculative; plausibility hinges on cited company positioning and market dynamics.
Sources
PUBLIC
[Automatika Robotics, 2026] Homepage | https://automatikarobotics.com/
[Freeway Ecosystem] Freeway Ecosystem | https://ecosystem.freewayphx.com/companies/automatika_robotics
[Techstars, 2026] Techstars unveils the 2023 Class of the Torino Cities of the Future Techstars Accelerator | https://www.techstars.com/newsroom/techstars-unveils-the-2023-class-of-the-torino-cities-of-the-future
[GitHub, 2026] Automatika Robotics GitHub | https://github.com/automatika-robotics
[Inria, 2026] Abdullah Haroon Rasheed - Morpheo | http://morpheo.inrialpes.fr/people/ahrasheed/
[LinkedIn] Automatika Robotics LinkedIn | https://fr.linkedin.com/company/automatika-robotics
[MarketsandMarkets, 2023] Robot Software Market Report | https://www.marketsandmarkets.com/Market-Reports/robot-software-market-202068448.html
Articles about Automatika Robotics
- Automatika Robotics Builds a Software Stack for the Physical AI Agent — The Inria spinout is developing an open-source ROS2 framework to give robots the ability to understand, remember, and act on context.