A foundational representation for an orchestrated lab

Self-driving laboratories (SDLs) promise accelerated scientific discovery and product development by closing the loop between robotic execution and AI/ML-driven decision making. In practice, however, SDL orchestration remains fragmented; workflows are typically encoded as laboratory-specific scripts or bespoke frameworks that are difficult to reproduce, maintain, or transfer across instruments and sites. As a result, many SDL implementations remain effectively one-off systems: hard to reproduce, costly to maintain, and difficult to extend beyond isolated case studies. This fragmentation is amplified by an artificial split between physical laboratory actions and the ``virtual'' operations that transform data into decisions, which are often treated as an external afterthought. Here, we propose a deliberately minimal, language-agnostic orchestration foundation that grounds a transferable workflow structure across both physical and computational domains. The framework organizes SDL behavior into four nested constructs (primitives, unit operations, state-preserving flows, and workflows) supporting validation, safer parallelism, and interoperability. We apply the hierarchy to four distinct SDL archetypes drawn from the literature illustrating those recurring orchestration patterns. Finally, we outline how the same typed, declarative foundation enables "Lab as Code": representing experimental environments and procedures as versioned, redeployable specifications ---analogous to Infrastructure as Code ---so workflows can be validated, shared and re-instantiated reproducibly across laboratories. We present this framework as a first iteration intended to evolve through community use, critique, and extension toward a shared foundation for scalable autonomous science.