1. Executive Overview {#executive-overview}

AICA Sàrl is a Swiss software company building an AI platform for adaptive robotics, headquartered at EPFL Innovation Park in Lausanne — one of Europe's most recognized deep-tech incubator environments. The company's core offering is a software platform that abstracts the complexity of robot integration and programming, enabling engineers to build advanced robotic applications across diverse hardware through an interactive, visual development environment. Its positioning at the intersection of AI, real-time control, and open extensibility places it squarely in the emerging category of robot software middleware and application platforms.

The company has attracted attention from industrial venture capital, with Momenta — a specialist mobility and industrial technology investor — noting AICA in its coverage as of March 2025. AICA's inclusion in Swiss startup delegation activity (startupticker.ch, VivaTech 2026) further signals that the company is considered a representative voice for Swiss deep-tech on international stages. These are meaningful third-party signals for an early-stage software robotics firm.

Not yet disclosed: founding date, headcount, total funding raised, and named customer deployments. Parties with accurate data are invited to claim or correct the record via the appropriate channel.

2. The Company Story {#the-company-story}

AICA Sàrl is legally registered as a Swiss limited liability company (Société à responsabilité limitée) and operates from EPFL Innovation Park in Lausanne — an address that carries deliberate signal. EPFL (École Polytechnique Fédérale de Lausanne) is one of Europe's foremost technical universities, with particular strength in robotics, machine learning, and control systems. Companies that establish themselves within its innovation park frequently maintain research adjacency to the institution, though AICA has not publicly confirmed any formal EPFL collaboration or spin-out status.

The company's self-described mission is to simplify robot integration and programming across diverse hardware, building toward more capable and flexible robotic systems. The name "AICA" and the tagline "The AI Platform for Adaptive Robotics" frame the company's value proposition around adaptability — a deliberate contrast to the historically rigid, hardware-specific programming paradigms that have dominated industrial robotics. The platform approach (SDK, open extensibility, growing component library) suggests a strategy of building an ecosystem rather than a closed product.

Not yet disclosed: the precise founding year, founding team composition, and the trajectory of fundraising rounds. Momenta VC's coverage (March 2025) represents the clearest external milestone in the public record to date. The company is actively recruiting a Business Development Manager/Sales Engineer based in Lausanne, which is consistent with a company transitioning from product development toward scaled commercial deployment.

3. Product Portfolio {#product-portfolio}

Products & versions

C30

DELIVERY

DINERBOT T9 Pro

HOSPITALITY

KLEENBOT C30

CLEANING

BUTLERBOT W3

DELIVERY

DINERBOT T9

DELIVERY

T10

HOSPITALITY$15,000

AICA's product portfolio, as publicly described, centers on a single cohesive software platform rather than a family of discrete hardware or hardware-bundled products. The flagship offering — referred to as AICA and specifically surfaced through AICA Studio — is a software environment for robot integration and programming. Its defining features include an interactive drag-and-drop graph editor for building robotic applications visually, real-time control and monitoring capabilities (covering position, velocity, and force controllers), a motion generation and reinforcement learning library, and an open SDK that allows developers to extend the platform's functionality with custom components.

The platform's design philosophy is hardware-agnostic: it connects to industry-standard robot arms rather than being tied to any single vendor's ecosystem. This positions AICA as a layer of software intelligence and integration that can, in principle, run across multiple robotic hardware platforms — a significant differentiator in a market where most robot programming toolchains remain proprietary to specific OEMs. The "growing library of smart components" language suggests an ongoing product development effort to expand the platform's capabilities over time.

A second product entry labeled "japan-2025" appears in extracted site data but lacks sufficient public detail for substantive analysis at this time. Not yet disclosed: whether this represents a market-specific product variant, a partnership announcement, or a forthcoming release. AICA is invited to clarify.

4. Technology Stack {#technology-stack}

Based on the publicly described key features of the AICA platform, several technology choices and architectural decisions can be identified or reasonably inferred.

The drag-and-drop graph editor paradigm for building robotic applications is consistent with a dataflow or component-based programming model — an approach also used in frameworks such as ROS 2 (Robot Operating System) node graphs and similar visual programming environments. Our read: AICA likely builds on or is philosophically aligned with component-based robotics middleware, where discrete software "components" are wired together to form application logic, rather than requiring monolithic code authorship.

The explicit mention of real-time control with position, velocity, and force controllers indicates that the platform operates at or near the hardware control loop level — typically requiring deterministic execution and low-latency communication. Our read: this is non-trivial engineering; real-time robotics control imposes strict requirements on operating system scheduling and communication protocols (e.g., EtherCAT, real-time Linux kernels). The fact that AICA names force control specifically suggests capability relevant to contact-rich manipulation tasks, not merely pick-and-place.

The inclusion of a reinforcement learning library alongside motion generation components is notable. Our read: this suggests the platform aims to close the gap between classical trajectory-based robot programming and learned, adaptive behaviors — which aligns directly with the "adaptive robotics" positioning. Whether this is inference-time deployment of pre-trained policies, online learning, or both is not yet disclosed.

The open SDK for extending functionality is a deliberate architectural choice signaling an ecosystem strategy: AICA appears to be building toward a marketplace or library of third-party components, not merely a closed toolkit. Limited public technical detail is available on the underlying languages, supported robot communication protocols, or cloud/edge deployment architecture.

5. Research, Papers, Authors, Labs {#research-papers}

AICA does not, based on available public data, present itself as a research-publishing organization. This is entirely consistent with its profile as a commercial software platform company. The company's adjacency to EPFL may mean that individuals on the team have prior academic publication records, but no papers, authors, or laboratory affiliations are surfaced in AICA's own public materials. Not yet disclosed: any formal research partnerships, academic co-authorship, or EPFL affiliation status. AICA is invited to surface relevant research credentials if applicable.

6. Media Evidence {#media-evidence}

Three independent third-party sources appear in the public record. Startupticker.ch — Switzerland's primary startup news platform — lists AICA among the Swiss delegation preparing for VivaTech 2026, a major European technology and innovation conference held in Paris. This signals institutional recognition within the Swiss startup ecosystem. Momenta VC (momenta.vc), a venture capital firm specializing in industrial technology and mobility, published coverage referencing AICA in March 2025, providing external investor-community validation. AI-booster.ch included AICA in blog coverage dated May 2025, indicating ongoing visibility within Swiss AI and technology media. No major international robotics trade press coverage (e.g., The Robot Report, IEEE Spectrum, TechCrunch) has been identified in the current data set.

7. Commercial Reality {#commercial-reality}

Revenue, annual recurring revenue, customer count, and specific deployment ROI figures are not disclosed in any public source reviewed. These metrics are rendered here as Not disclosed. Companies with accurate commercial data — including AICA itself — are invited to claim, correct, or supplement this record through the appropriate submission channel.

What can be stated from public signals: the active recruitment of a Business Development Manager/Sales Engineer in Lausanne is consistent with a company that has achieved product maturity sufficient to warrant a dedicated sales function, but has not yet scaled a large commercial sales team. The Momenta VC reference (March 2025) may indicate a funding event or active fundraising process, though no round size or stage has been confirmed publicly. Our read: AICA is likely in an early-to-growth commercial phase — past initial product development, moving toward repeatable customer acquisition.

8. Markets and Use Cases {#markets-use-cases}

AICA's product descriptions do not specify named industry verticals or tagged use cases in the current public data, but the platform's technical characteristics clearly imply target markets.

The combination of position, velocity, and force controllers with motion generation points most directly to industrial robotic manipulation — tasks such as assembly, machine tending, quality inspection, and surface finishing, where force-compliant motion and flexible programming are competitive advantages over rigid, teach-pendant-based approaches. The hardware-agnostic design broadens the addressable base across any facility operating standard industrial robot arms (6-axis collaborative or industrial robots from major OEM brands).

The reinforcement learning library opens a secondary addressable market: applications where robot behavior must adapt to variation in parts, environments, or tasks — scenarios that classical programming handles poorly. This includes flexible manufacturing, small-batch production, and potentially research and development environments where novel manipulation tasks are prototyped. The open SDK further suggests appeal to system integrators and robotics OEMs seeking to embed adaptive software capabilities into their own offerings without building the underlying platform from scratch.

AICA's presence at events such as VivaTech and in Swiss startup channels also suggests some orientation toward European industrial automation markets, where regulatory familiarity and local presence are valued by manufacturing customers.

9. Competitive Landscape {#competitive-landscape}

The robot software platform and middleware category has grown substantially as the industry has recognized that hardware commoditization shifts value toward software. AICA competes — or will compete — in a space that includes both established automation software vendors and a cohort of well-funded robotics software startups building visual programming, real-time control, and AI-augmented motion planning tools. The hardware-agnostic, SDK-open, AI-inclusive positioning is a credible strategic stance within this category.

What distinguishes or could distinguish AICA is the combination of real-time force control depth with a visually accessible development environment and a reinforcement learning component library — a stack that attempts to serve both experienced robotics engineers and organizations seeking to lower the barrier to advanced robot programming. How this differentiation holds as the category matures will depend on ecosystem growth (component library breadth, SDK adoption) and validated customer deployments. The module above provides current competitive context.

10. Country Advantage / Geopolitical {#geopolitical}

Switzerland's position as a non-EU but deeply integrated European economy provides AICA with a specific set of advantages relevant to industrial robotics software. Swiss companies benefit from strong intellectual property protections, access to European markets via bilateral agreements, and a reputation for precision engineering that carries meaningful brand weight with European manufacturing customers. EPFL's international standing in robotics and machine learning research also provides AICA with a talent pipeline and credibility signal that few other geographic locations can match.

For an AI software company, Switzerland's evolving regulatory posture toward AI — distinct from but observant of the EU AI Act — may become a relevant differentiator as industrial customers in regulated sectors assess software vendor risk. Our read: AICA's Swiss domicile and EPFL adjacency are genuine assets in European enterprise sales contexts, where provenance, data sovereignty, and regulatory clarity are increasingly procurement criteria.

11. Hype vs Real vs Ugly {#hype-real-ugly}

Real (verified from public sources): AICA operates a software platform for robot integration and programming with a documented feature set including a drag-and-drop graph editor, real-time position/velocity/force controllers, a motion generation and reinforcement learning library, and an open SDK. The company is physically located at EPFL Innovation Park in Lausanne. Third-party sources — Momenta VC, startupticker.ch, and ai-booster.ch — independently reference the company in 2025.

Company claims (stated by AICA; not independently verified here): The platform "simplifies robot integration and programming across diverse hardware." AICA Studio enables users to "build advanced robotic applications" with a drag-and-drop interface. The component library is described as "growing." These are product claims made on the company's own site and should be evaluated against demonstrated customer deployments when those become available.

Gaps requiring disclosure: Not yet disclosed: customer names, deployment scale, performance benchmarks for the RL library, supported robot arm hardware compatibility list, pricing model, and funding history. The "japan-2025" product/entry lacks any public description and cannot be assessed. AICA is invited to address these gaps to support a more complete picture for prospective customers and partners.

Our read: There is no evidence of materially misleading claims in AICA's public positioning. The platform description is technically coherent and the company's third-party footprint, while modest, is growing. The primary risk is the gap between platform capability claims and publicly demonstrated commercial traction — a gap common to early-stage B2B software companies and not in itself a negative signal.

12. Future Scenarios {#future-scenarios}

Our read — Bull case: AICA establishes itself as a preferred software layer for flexible robotic cell integration in European manufacturing, driven by ecosystem growth (component library, SDK adoption by system integrators), a successful commercial scaling led by the incoming BD/Sales hire, and potential formal partnership or spin-out recognition from EPFL. Momenta VC's involvement, if it represents an investment, could accelerate both funding and industrial customer introductions. The reinforcement learning library becomes a meaningful differentiator as adaptive manufacturing demand grows.

Our read — Base case: AICA grows steadily within a niche of technically sophisticated robotics system integrators and OEM partners who value the platform's openness and real-time control depth. Commercial scale remains moderate through 2026, with the company building a reference customer base in European industrial automation. Ecosystem growth is gradual but consistent. The company remains an EPFL Innovation Park tenant and continues to surface in Swiss startup channels as a recognized but not yet widely deployed platform.

Our read — Bear case: The robot software middleware market consolidates around better-capitalized platforms or established automation OEMs who build comparable visual programming and AI tooling internally. AICA's hardware-agnostic positioning, while strategically sound, requires sustained ecosystem investment that is difficult to finance at early stage. Talent competition in the Lausanne/Zurich corridor is intense, and the sales cycle for industrial robotics software is long. Without a disclosed funding base or named customer anchor, growth stalls at the pilot/evaluation stage.

13. What to Watch {#what-to-watch}

• Funding announcement: Any disclosed investment round (size, lead investor, stage) will be the single most informative commercial signal. Momenta VC's March 2025 coverage warrants monitoring for a formal announcement.
• Named customer deployments: First publicly named industrial customer or system integrator partnership will validate the platform's commercial readiness and target vertical.
• SDK and component ecosystem growth: Watch for developer documentation releases, component library expansion announcements, or third-party integrator endorsements — these are leading indicators of platform strategy execution.
• "japan-2025" entry clarification: This unresolved product/entry may represent a market entry, partnership, or product launch in Japan — a significant industrial robotics market. Any clarification would materially update the company's geographic and commercial picture.
• EPFL affiliation disclosure: Formal confirmation or denial of EPFL spin-out status or research collaboration would inform both technology credibility and IP ownership assessments.
• VivaTech 2026 participation: AICA's presence in the Swiss delegation for VivaTech 2026 (per startupticker.ch) offers a near-term public milestone for product demonstrations and partnership announcements.
• BD/Sales hire and hiring trajectory: Progress on the Business Development Manager/Sales Engineer role, and any subsequent hiring, will signal commercial scaling pace.

14. Sources & Methodology {#sources-methodology}

Primary data source: Content extracted from AICA's own public website (aica.tech), including structured schema data, product descriptions, key feature lists, and careers page copy. All content derived from this source is treated as company-claim and labeled accordingly throughout this report. It represents the company's own characterization of its products, mission, and positioning — not independently verified technical or commercial fact.

Third-party press sources (independent): Three external sources were identified and cited by outlet name: startupticker.ch (Swiss startup media), momenta.vc (industrial technology venture capital), and ai-booster.ch (Swiss AI media). These are treated as independent third-party signals, cited with dates where available, and not conflated with company claims.

Methodology rubric (applied uniformly to every company assessed under this framework):

1. Factual claims are grounded exclusively in the data provided — no external knowledge, invented specifications, or assumed partnerships are introduced.
2. Negative characterizations are expressed only as fixable gaps with an invitation to correct, as labeled inferences ("Our read:"), or as labeled company claims under scrutiny.
3. Sections lead with verified strengths before gaps.
4. Live data modules (products, news, papers, media, customers, competitors, claim-tracker) are placed exactly as specified; prose is kept appropriately brief at those insertion points.
5. Inferences are explicitly flagged; analyst judgment is separated from verified fact throughout.