Software catalogue

All the CoreSense software is open source and hosted in the CoreSense GitHub organisation. This page lists the main repositories, grouped by topic. The Install column shows how each one can be installed (see Build and Install): Pixi, apt or source.

At the end of the project, on 30 September 2026, 80 packages of 20 repositories were added to the CoreSense Pixi channels, for ROS 2 Jazzy and Kilted, together with the dependencies that were not available in RoboStack (behaviortree_ros2, nlohmann_json_schema_validator_vendor, launch_pal, llama_msgs and kb_msgs, the latter from a CoreSense fork that builds on Jazzy and Kilted). When only some packages of a repository are in the channels, the Install column names them.

Architecture

Repository	Description	Install
cs4home_architecture	ROS 2 implementation of the CoreSense architecture: cognitive modules (afferent, core, efferent, meta and coupling components) and flows.	Pixi (`cs4home_core`), source
cs4home_examples	Example cognitive modules built with `cs4home_architecture`.	source
cs_functional_module_template	Template to start a new functional module.	source
coresense_understanding	Understanding system: generates strategies to obtain models with given properties.	Pixi, source
understanding-logic	Logic of the understanding core.	source
coresense_understanding_examples	Example engines for the understanding system.	source
decision_system	CoreSense decision system for ROS 2.	Pixi, source
coresense_common	Common messages, bringup and behaviour-tree controller of the understanding system.	Pixi, source
coresense_engine_examples	Examples of CoreSense engine nodes.	Pixi, source
triplestar_kb	Knowledge base for ROS 2 based on RDF-star and RDF 1.2.	Pixi, source
knowledge_core	RDFlib-backed minimalistic knowledge base with ROS 2 API, OWL2 reasoning and event subscriptions. Used as the per-drone KB in the inspection testbed.	Pixi, source
coresense_vampire	ROS 2 node that wraps the Vampire automated theorem prover.	Pixi, source
cso	CoreSense Ontology, resolvable at https://w3id.org/coresense/cso.	–

Cognitive modules and structures

Repository	Description	Install
EasyNavigation and easynav_plugins	EasyNav, a plugin-based, real-time navigation framework able to include semantic and awareness representations. Documentation: https://easynavigation.github.io.	Pixi, apt
risk-awareness-module	Risk awareness module (RiskAM): real-time risk score of visually navigated robots. See Risk awareness with RiskAM.	Pixi, source
physics-aware-module	Physics-aware modelling module based on neuro-evolutionary symbolic regression.	Pixi, source
fms	Cognitive structure to deploy AI foundation models inside CoreSense systems.	source
cs-ne	CoreSense Navigation Essential.	source
cs4home_sound_module	Cognitive module for sound perception.	Pixi (`sound_msgs`), source
cs4home_vision_module	Cognitive module for visual perception.	Pixi (`cs4home_msgs`), source
cs4home_person_tracker_module	Person tracking from camera and laser detections.	source
cs4home-explainability	Explainability framework based on behaviour-tree status and component evidence.	Pixi, source
coresense-explainability	Explainability framework templates: example component explainers and explainer selector, to start custom explainers.	Pixi (the three example explainers), source

Toolchain

Repository	Description	Install
rossdl	ROS System Definition Language: model-based description of ROS 2 systems and code generation.	Pixi (`rossdl_cmake`), source
RosTooling	Eclipse-based IDE to model ROS systems. See CoreSense Toolchain.	update site
CS_ros2model_TBs	Models of the testbed systems extracted with the introspection tools.	–
SysML2-server-api	Dockerised local SysML v2 server and API pipeline.	Docker

ROS 2 instrumentation

Repository	Description	Install
coresense_instrumentation	Virtual drivers to activate, deactivate and monitor data flows, and an RViz plugin. See CoreSense Instrumentation.	Pixi, source

Testbeds

Repository	Description	Install
aerostack2	CoreSense fork of Aerostack2, the ROS 2 framework for autonomous aerial systems used as the runtime platform in the inspection testbed, with the CoreSense behaviours (auction-based task allocation, collision avoidance, collective awareness). See Aerostack2 documentation.	Pixi, source
CoreSense4Home	CoreSense implementation for RoboCup@Home (social testbed).	source
cs4home_hri_challenge	HRI challenge of the social testbed.	source
tiago_sw	Open-source software for the PAL TIAGo robot used in the social testbed.	source, apt
collective_awareness_structure	Collective awareness for multi-robot systems built with Aerostack2 (inspection testbed).	Pixi, source
tb2_project	Simulation of the drone panel inspection testbed.	source