THE EXPERIENCE DESIGN INSTITUTE

Building the Infrastructure for Human-Robot Trust.

The research project behind LSEP — the open standard for how robots communicate intent to humans. We maintain the open LSEP standard for human-robot safety signalling.

The Experience Design Institute (EDI) is the research project that stewards LSEP (Luminae Signal Expression Protocol), an open, MIT-licensed standard for how robots signal state and intent to people through light, sound, and motion. Designed with the EU AI Act (Annex IV) and ISO 13482 in mind; conformance is not claimed.
Open Standard
MIT-Licensed
Physics-Based
Non-Discriminatory
THE LSEP MANIFESTO

What We Believe

Our Belief: Silence between humans and autonomous machines is dangerous. Opaque robot behavior breeds fear and resistance. When a 200kg machine approaches you and you have no idea what it's about to do — that's not innovation. That's negligence.

Our Mission: We believe every human has the fundamental right to know what a physical AI system is about to do. Not through screens. Not through apps. Through the machine itself — directly, physically, unmistakably. This is the foundation of human-robot interaction.

Our Solution: We don't build robots. We build the open standard for trust between humans and machines. Physics-based, universally understood, and designed for interoperability across every platform. No discrimination, no surveillance. Just light, sound, and motion.

Our Standard: LSEP is not a product. It's infrastructure. Like Bluetooth gave us wireless audio, like USB-C gave us universal power — LSEP gives robots a voice. An honest one. Built for robot safety, designed for transparency.

THE PROTOCOL

1 Protocol. 9 States. 3 Modalities.

LSEP (Luminae Signal Expression Protocol) is a formal Human-Robot Interaction (HRI) standard for intent signaling and algorithmic transparency. A robot communication protocol independent of facial recognition. The robot's perception stays private; its intent becomes visible. Every signal is physics-based. Every state is human-legible. Every implementation is auditable and built for interoperability across platforms.

Light

LED arrays and light patterns communicate intent. Color-coded states: red for care and critical states, green for safe approach, yellow for uncertain motion. Visible from all angles, impossible to miss.

Sound

Auditory signals provide redundancy. Frequency-based notifications allow hearing-impaired users to understand robot state. Sub-audible warnings for precision navigation.

Motion

Kinematic expressions — predictable movement patterns that signal intent. Deceleration, orientation, and temporal pacing all convey meaning. Physics-based and learnable.

9
Protocol States
3
Signal Modalities
1
Open Standard
LSEP SYSTEMS ARCHITECT PROGRAM

Learn the standard. Prove your implementation.

The LSEP Systems Architect Program trains engineers and safety teams to implement the standard, with a structured assessment at the end. Assessed implementations are documented, auditable, and human-legible. Conformance with the EU AI Act or ISO 13482 is not claimed.

For Manufacturers

Implementing LSEP signals commitment to human-centered design and interoperability. A documented, auditable signalling layer is useful groundwork for regulatory work. It does not replace conformity assessment.

For Safety Officers

Standardized, auditable HRI. Robots implementing LSEP use one consistent signal vocabulary, designed to reduce confusion about machine behavior. Full protocol transparency for internal auditing.

For Insurers

A shared signalling standard makes human-robot interaction observable and comparable across platforms. We see potential for benchmarking and risk assessment. Outcome data does not exist yet, so we make no outcome claims.

TRAINING

Master the Standard.

Our training program guides manufacturers, engineers, and safety teams through LSEP integration. Learn the 9 protocol states, implement the 3 signal modalities, and achieve interoperability across platforms. Hands-on workshops. Real-world case studies. Ongoing support.

GOVERNANCE

The Team Behind the Standard.

A Swiss engineering and research team building the infrastructure layer for human-robot trust.

Nemanja Galić NG

Nemanja Galić

Founder & Experience Designer

Experience Designer and researcher at the intersection of human cognition, physical AI, and trust engineering. Creator of the LSEP protocol. Based in Bern, Switzerland.

Mijo Barbaric MB

Mijo Barbaric

IT Architecture Advisor

Advises on enterprise scalability, cloud architecture, and deployment across diverse robotic platforms.

Michael Aebli MA

Michael Aebli

Software Engineering Advisor

Brings deep software engineering expertise to LSEP's reference implementation and developer tooling ecosystem.

THE LSEP ALLIANCE

Join the companies building the future of HRI.

The LSEP Alliance connects manufacturers, integrators, research institutions, and safety bodies. Collaborate on standards, share knowledge, and accelerate human-robot interaction adoption across industries.

Network Access

Connect with leading robotics manufacturers, systems integrators, and safety organizations. Build partnerships. Share solutions. Accelerate market adoption.

Technical Resources

Access reference implementations, developer tools, and API documentation. Contribute to the open standard. Shape the future of human-robot interaction.

Industry Recognition

Be listed as an official LSEP Alliance member. Gain credibility with regulators, insurers, and customers. Demonstrate commitment to transparent, safe HRI.

FREQUENTLY ASKED QUESTIONS

Technical FAQ

What is LSEP?
LSEP (Luminae Signal Expression Protocol) is an open-source communication standard that enables robots and autonomous machines to communicate their internal states and intentions to humans through physics-based signals — light patterns, sound frequencies, and kinematic expressions. It is stewarded by the Experience Design Institute (EDI).
How does LSEP relate to the EU AI Act?
LSEP is designed with the EU AI Act in mind. Protocol-level state logging and human-legible intent signalling map onto the Act's transparency themes (Article 13; Annex IV documentation for high-risk systems). Implementing LSEP does not by itself make a system compliant; conformance is not claimed.
Is LSEP compatible with ROS 2?
Yes. LSEP is designed as a middleware-agnostic communication layer. Reference implementations are available for ROS 2 (Robot Operating System), and the protocol can be integrated with any robotics framework that supports structured state messaging.
What robot types does LSEP support?
LSEP supports humanoid robots, collaborative robots (cobots), autonomous mobile robots (AMRs), industrial manipulators, and service robots. The protocol is platform-independent and scales from single-robot deployments to multi-robot fleet operations.
How does LSEP support interoperability?
LSEP defines a universal state machine with 9 standardized states and 3 signal modalities. Any robot implementing LSEP uses the same signal vocabulary, so state signalling stays consistent across manufacturers and platforms.
Where can I find the LSEP specification?
The full LSEP specification is available on lsep.org and the GitHub repository at github.com/NemanjaGalic/LSEP. The specification is open-source and free to implement.