Product profile
AT-SYNAPSE
AT-SYNAPSE connects MCP-compatible LLM agents with ROS 2 robot functions. The platform accepts voice or text instructions, interprets camera and LiDAR data...
Facts
- Company
- AttracLab
- Product type
- robot_ai_platform
- Status
- development
- Primary capability
- LLM-to-robot orchestration
- Website
- https://prtimes.jp/main/html/rd/p/000000107.000052796.html
About AT-SYNAPSE
AT-SYNAPSE is AttracLab's robot-AI platform connecting MCP-compatible language-model agents with ROS 2 functions. It interprets voice or text instructions and can call tools for vision, speech, motion, navigation and safety using camera and lidar data.
The software is designed for service robots, AMRs, UGVs and research platforms. Supported behaviour depends on the connected model, available ROS 2 tools, robot hardware and safety configuration; AttracLab introduced AT-SYNAPSE in July 2026 as a development-stage system.
Market context
ROS 2, robot_agent_orchestration, Model Context Protocol, official_architecture_diagram_fitted_to_plain_white_16_9_canvas, https://prcdn.freetls.fastly.net/release_image/52796/107/52796-107-69b84ace25c7d2e45d6ec7821d81ecd9-1536x1024.jpg, software_stack, developer_platform, 2026-08-21, Developer Platform, developer_platform, developer_platform, controlled_category_mapping_2026-08-21, false
Summary
Type: Robot AI platform
Core: AT-SYNAPSE
Function: Connect systems
Audience: Robot developers
Specifications
- agent_interface
- Model Context Protocol (MCP)
- console_connectivity
- rosbridge and WebSocket
- instruction_inputs
- Voice and text
- middleware
- ROS 2
- monitoring_interface
- AT-SYNAPSE Robot Console web interface
- safety_behavior
- Obstacle-triggered automatic stop and confirmation before ambiguous hazardous actions
- sensor_inputs
- Camera and 3D LiDAR
- tool_groups
- Vision, speech, face, motion, navigation and safety 6 groups
Operations
- Deployment environment
- robot development, service robots, mobile robots, field robotics
- Primary user
- Robot developers, integrators and field operators
- Workflow
- Orchestrate LLM-driven perception, dialogue, navigation, motion and safety tools
- Interaction model
- Voice or text commands with browser-based state and reasoning monitoring
- Mobility class
- Software platform
- Product detail
- ROS 2, robot_agent_orchestration, Model Context Protocol, official_architecture_diagram_fitted_to_plain_white_16_9_canvas, https://prcdn.freetls.fastly.net/release_image/52796/107/52796-107-69b84ace25c7d2e45d6ec7821d81ecd9-1536x1024.jpg, software_stack, developer_platform, 2026-08-21, Developer Platform, developer_platform, developer_platform, controlled_category_mapping_2026-08-21, false