Perception
Connect cameras, microphones, sensors and external perception models to a shared live interaction state.
IronHeart.AIpatent pendingCognitive runtime infrastructure for physical AI—integrating perception, real-time voice conversation, persistent memory, personality, neuroscience-based gaze coordination and actions.
Hardware, locomotion and task models do not create believable social interaction. IronHeart coordinates the signals between what a robot perceives, remembers, says, looks at and does.
Not another chatbot inside a robot. A continuous cognitive loop built for embodiment.Connect cameras, microphones, sensors and external perception models to a shared live interaction state.
Low-latency, interruptible conversation with natural turn-taking and multilingual behavior.
Carry user, session and entity context across interactions instead of resetting after every encounter.
Give each robot a stable identity, role, tone, behavioral boundaries and relationship continuity.
Coordinate attention, speech and gaze using neuroscience-informed interaction patterns.
Turn intent into physical or digital execution through robot controls, tools, workflows and APIs.

IronHeart has been developed through real robots, live conferences, education settings and noisy public environments—where latency, interruptions, attention and connectivity become product requirements.
In April 2026, Hidoba Research took part in a European premiere at Willms-Gymnasium in Delmenhorst, Germany: Captcha, a humanoid robot powered by IronHeart, conducted a simulation of the school’s Abitur oral examinations with real students.
This kind of work requires perception to follow the student, realtime voice to conduct the exchange, memory to preserve the exam thread, personality to maintain the examiner role, gaze coordination to direct attention and actions to ask follow-ups and produce a structured outcome.



At CES in Las Vegas, Captcha operated as an autonomous social robot in a live public environment—moving through the exhibition space, approaching visitors, introducing itself and holding spontaneous conversations.
The important behavior appeared between the conversations. When a visitor did not answer or ignored the approach, IronHeart treated the absence of engagement as a social signal. Captcha could adjust its timing, try a different prompt or disengage and continue through the space instead of repeating a fixed script.
Perception detects people and attention. Realtime voice opens and sustains the exchange. Memory preserves the encounter. Personality determines how Captcha introduces itself. Gaze signals who it is addressing. Actions connect the interaction to autonomous movement—one continuous Cognitive Runtime operating in the physical world.
Start with the managed API, move interaction-critical paths to the edge, or run inside private infrastructure. The runtime adapts to the product—not the other way around.
Prototype and launch connected products quickly.
Keep critical interaction available when networks are slow or unavailable.
Deploy inside customer-controlled infrastructure for sensitive environments.
IronHeart replaces one-size-fits-all behavior with a runtime configured for each embodiment, environment and commercial role.
Embed persistent personality, voice, memory and social behavior across a product line.
Configure knowledge, policies, actions and deployment for healthcare, education, retail, hospitality and public services.

We work alongside your engineering team from evaluation through fleet deployment.
interaction: "persistent"
perception: [vision, audio, proximity]
voice: { realtime: true, interruptions: true }
memory: { user: true, entity: true }
gaze: { coordination: "adaptive" }
actions: [gesture, navigate, tools]These plans cover the managed realtime API. Edge, offline, private cloud and OEM production licensing are scoped separately.
Early robot prototypes and API validation.
Customer-facing pilots and recurring deployments.
High-volume voice and robotics applications.
Architecture workshop, integration support, model and data controls, custom runtime configuration, fleet scaling and commercial licensing.
The cognitive layer can power different robot bodies, device classes and vertical products.
Memory, personality and coordinated behavior turn commodity intelligence into a durable product experience.
Cloud usage, production integration, edge licensing and private infrastructure serve different stages of the robotics market.
Tell us about the platform, use case and deployment environment. We’ll map the fastest route from evaluation to production.
[email protected]