Researchers compare four formalisms for programming robot missions

As robots take on increasingly complex roles across industries, specifying what those machines should do has become a critical challenge—yet there's no consensus on the best way to do it. A new research paper from arXiv surveys four major formalisms that robotics teams use to define missions, evaluating each approach's strengths and limitations.

The challenge of mission specification#

Modern robots must operate flexibly across diverse environments and handle multiple tasks. The people who need to program these missions are often domain experts—field operators, engineers, supervisors—rather than software specialists. They need tools that are intuitive, expressive enough to capture real-world complexity, and well-supported by existing software platforms. Currently, there is no standard method for mission specification in single-robot or multi-robot systems.

Four competing approaches#

The researchers analyzed four widely-used formalisms:

• Behavior Trees: Structure missions as hierarchical trees of behaviors, each with success/failure logic
• State Machines: Represent missions as sets of states and transitions between them
• Hierarchical Task Networks (HTN): Break missions into abstract tasks that decompose into concrete subtasks
• Business Process Model and Notation (BPMN): Originally developed for business workflows, increasingly adapted for robotics

Each formalism offers different levels of abstraction, expressiveness, and integration capabilities with human processes and external devices.

What the analysis covers#

The study moves beyond general software engineering to focus specifically on mission-level descriptions for robots. The researchers examined each formalism's underlying control structures, evaluated how well they model real-world robotic scenarios, and assessed the practical tool support available to practitioners. They also validated their conclusions with input from domain experts.

Why it matters#

Understanding the fit between a formalism and actual robotic needs could help teams select the right approach for their system. By mapping out what each formalism does well and where it falls short, the work aims to guide both practitioners designing robot systems and researchers developing better mission-specification tools.