Toward an Agricultural Operational Design Domain: A Framework

TL;DR

This paper defines the Agricultural ODD (Ag-ODD) Framework to address the unique challenges of autonomous farming in dynamic, off-road environments. It extends the ASAM OpenODD foundation with a CityGML-inspired context, and augments the PEGASUS-based 7-Layer Model with a new process layer to capture agricultural work, enabling iterative verification against logical scenarios. The framework is demonstrated through cultivation and wheat-harvesting use cases, illustrating how use cases drive Ag-ODD derivation, scenario generation, and verifiable boundaries. By integrating standardized environmental description, process awareness, and rigorous testing, the Ag-ODD Framework aims to standardize and scale environmental descriptions, facilitate simulation-based validation, and support safety compliance for autonomous agricultural systems.

Abstract

The agricultural sector increasingly relies on autonomous systems that operate in complex and variable environments. Unlike on-road applications, agricultural automation integrates driving and working processes, each of which imposes distinct operational constraints. Handling this complexity and ensuring consistency throughout the development and validation processes requires a structured, transparent, and verified description of the environment. However, existing Operational Design Domain (ODD) concepts do not yet address the unique challenges of agricultural applications. Therefore, this work introduces the Agricultural ODD (Ag-ODD) Framework, which can be used to describe and verify the operational boundaries of autonomous agricultural systems. The Ag-ODD Framework consists of three core elements. First, the Ag-ODD description concept, which provides a structured method for unambiguously defining environmental and operational parameters using concepts from ASAM Open ODD and CityGML. Second, the 7-Layer Model derived from the PEGASUS 6-Layer Model, has been extended to include a process layer to capture dynamic agricultural operations. Third, the iterative verification process verifies the Ag-ODD against its corresponding logical scenarios, derived from the 7-Layer Model, to ensure the Ag-ODD's completeness and consistency. Together, these elements provide a consistent approach for creating unambiguous and verifiable Ag-ODD. Demonstrative use cases show how the Ag-ODD Framework can support the standardization and scalability of environmental descriptions for autonomous agricultural systems.

Figures (7)

• Figure 1: Multiple dimensions of automation in agriculture: The individual degree of automation can be defined both for the driving task according to SAEJ30162021 and for the working process according to ISO18497 of the machine or device.
• Figure 2: The Ag-ODD Framework for deriving the agricultural operational design domain (Ag-ODD). The initial Ag-ODD and its associated logical scenarios are derived from the defined use cases . Once these are established, the iterative verification process begins. During this process, inconsistencies and gaps within the Ag-ODD definition are often exposed by the logical scenarios. Any resulting modifications must then be formalized as a revised Ag-ODD by comparing them to the initial input parameters. This iterative procedure, depicted by the two verification arrows, continues until the Ag-ODD reaches a stable state. Comparing the Ag-ODD with the input parameters: I) functional requirements, II) system capabilities, and III) the results of the Hazard Analysis and Risk Assessment (HARA) as detailed in \ref{['fig:Ag-ODD']}, as well as the logical scenario derivation as illustrated in \ref{['fig:Ag-PEGASUS']}.
• Figure 3: The Agricultural Operational Design Domain (Ag-ODD) concept. The derivation is done from the following three values: I) functional requirements, II) system capabilities, and III) the results of the Hazard Analysis and Risk Assessment (HARA) as framing limitation. The Ag-ODD is composed of four primary categories: the , the , the condition, and the . In addition, each attribute is defined by a and a . In order to ensure a comprehensive understanding, it is necessary to consider the categories of the operational context sequentially. The process category includes condition-dependent variables. The framing limitations III) HARA are influenced by the I) functional requirements and II) system capabilities , creating a cyclical dependency between the Ag-ODD and the framing limitations. The starting point for defining Ag-ODD is the use case or use cases in connection with the framing limitations or Ag-ODD itself.
• Figure 4: The 7-Layer Model is designed for use in agricultural scenarios and can be used to derive logical scenarios, including agricultural processes from use cases. The 7-Layer Model comprises the six PEGASUS layers PPO2021 with slight modifications, as well as a 7th process layer that can potentially alter the other layers through process descriptions.
• Figure 5: Iterative verification process between the Ag-ODD and the logical scenarios . A predefined Ag-ODD, as drawn as dark blue hexagon, means that the Ag-ODD is not yet well enough defined or that there are not yet enough scenarios. As soon as the scenarios cover the entire Ag-ODD, the Ag-ODD is verified against the scenarios and vice versa. A green dashed border means that this boundary is permissive ($\cup$)---everything that is not explicitly mentioned is still included; a red border means that this boundary is restrictive ($\cap$)---everything that is not explicitly mentioned is excluded. In the first two iterations, additional scenarios are added to better cover the Ag-ODD. Within the third iteration, the Ag-ODD is adjusted so that no further scenarios are need or because these parts of the Ag-ODD cannot be supported by framing limitations.