Senior Symoblic AI & Knowledge Engineer - Robot reasoning

Common Sense Robotics is a young startup based in Leuven, Belgium.

We build robot systems for "touch labour" challenges in manufacturing industries where the cost of failure is high, such as aerospace, automotive and construction. We focus on industries that expect every robot decision to be traceable, auditable, and explainable.

We now need extra brains to develop our "Task Execution System": a white-box, ontology-driven robotic skill stack designed from the ground up as a fully explainable "foundation model" for task specification, world modelling, skill generation, motion control and task-directed active perception.

While we might rely in some components of our skill stack on opaque learned policies, the backbone of our stack are ontologically represented work instructions, and task execution dependencies on the available resources (compute, communicate, move, perceive and reason). The formal models of tasks are automatically translated into orchestration and configuration actions on a rich repository of touch labour motion primitives. Those primitives are built on the solid foundations of advanced control theory of force sensing and real-time computer vision for robotic manipulation, exploiting state of the art that in some cases dates back already five decades.

That approach allows us to integrate deep and reinforcement learning policies with symbolic reasoning, to realise high-performance touch labour task executions that are fully predictable, inspectable, traceable and explainable.

The Role:

You will contribute to the reasoning layer of TES: the component that takes a queryable knowledge representation of a touch labour task, together with representations of available robot skills, and with access to a semantically labeled model of the actual status of the world around the robot, and turns that knowledge and information into an orchestrated set of concurrent robot activities. The reasoning is not activated just once, off line, before starting the skill executions, but it must run continuously alongside the robotic skills, dialoguing with them to assess task progress, to adapt their parameters, to start and stop activities, and to integrate learned components wherever appropriate.

So, offline the reasoner resolves task dependencies, sequences operations, parametrizes primitives against the ontology, and online it produces action commands and execution traces that downstream layers can run and that auditors can inspect.

Concretely, you will:

• Contribute to the symbolic reasoning core: improve and extend the task specification ontology, develop tooling to formulate queries and to solve those queries, resulting in the generation of a software architecture that connects to the touch labour control stack on the robots.
• Help define how robot capabilities, constraints, and task structures are represented, and how plans are derived, parametrized, and validated against.
• Develop the runtime reasoning activities that track task progress against symbolic models, adapt parameters online, trigger activity transitions, and integrate reinforcement learning where it fits within the symbolic frame.
• Build code generation from formal models so that symbolic specifications produce executable robot behaviour.
• Integrate your work with that of your colleagues responsible for the motion primitives repository, the automatic code generation tools, the multi-agent and realtime software architectures, and the development of applications and products.

Useful expertise:

• using Linux and other industry-grade open source software in all of your professional activities.
• Git, C/C++, Python.
• labeled property graph software, Graph Query Language, SPARQL, JSON-LD, RDF.
• writing documentation in semantic HTML

If this matches your expertise and ambitions, we look forward to your application.

How to apply: We don't review generic CVs. Instead, send us a short application document, written specifically for this role, explaining why you're a good fit and proposing concretely how you would approach one or two of the challenges described above. email: [email protected]