Behavior taxonomy, temporality, episode boundaries, and cross-embodiment mapping

Determine how to construct behavioral taxonomies, represent temporality, determine the boundaries of Behavior Episodes, and map equivalent behaviors across radically different robot and agent embodiments within the Behaviora architecture.

Background

Behaviora represents internal and external behavior as addressable Behavior Episodes composed of ordered components identified by persistent Internet of Behaviors (IoB) Addresses. The architecture depends on a meaningful behavioral vocabulary, but the paper does not provide a complete solution for organizing behavioral concepts into taxonomies, handling their temporal structure, or deciding where one episode ends and another begins.

The architecture is intended to preserve behavioral identity across robots and agents with different morphologies, sensors, actuators, and control systems. Consequently, mapping behaviors that are equivalent at the representational level but realized differently across radically different embodiments remains an explicit unresolved problem.

References

We are aware of several open questions, including how to construct behavioral taxonomies, temporality, how to determine Behavior Episode boundaries, and how to map equivalent behaviors across radically different embodiments.

— Behaviora - A Conceptual Architecture for External and Internal Behavior of Robots and Agents  (2609.27536 - Nyman, 23 Sep 2026) in Section 1, Introduction

We have not developed the IoS&P concepts yet and have left the taxonomy, identification system, and implementation of this layer open.

— Behaviora - A Conceptual Architecture for External and Internal Behavior of Robots and Agents  (2609.27536 - Nyman, 23 Sep 2026) in Section 2.6, External Behavior, Internal Behavior, and the excluded IoS&P Layer