Standarization

HRIM is an information model aiming to cover the current necessities in the robot/cobot industry. As such, reviewing it with manufacturers and main players in the industrial landscape is a must. Part of this is making sure we respect the different standards in play.

HRIM's development is closely tied to the guidelines defined in the (still in development) Robotics — Modularity for service robots — Part 1: General requirements (ISO Standard No. 22166) as our group participates in its development. That modularity vision has been at the heart of HRIM since the project's start.

Our review on OPC UA's current state shows that our structures are compatible and that most of what they consider necessary information is already contemplated in HRIM. Filling the gaps wouldn't entail re-structuring.

OPC UA is the soon-to-be de-facto standard for anything industrial. We consider it a must to be compatible with it (to some extent). Furthermore, checking our alignment with this standard serves as a way for us to validate our work.

As OPC UA is still in development, this alignment check will continue following their updates. Currently, this alignment has been checked against the publicly available OPC UA Companion Specification for Robotics: Part 1 draft.

For starters, the draft focuses on a structure to enable a description and/or "snapshot" (a representation of the system at a given time) of a robotic system for its management and maintenance. Because of this, the structures defined on it and the data they represent should be achievable through our information model, but don't necessarily imply a direct implementation of said structures. In Layman's terms, even if we don't contemplate information communication following the same structure OPC UA defines, an operator should be able to access all the information needed for an aforementioned description or "snapshot" solely through our information model. This means every mandatory piece of data should be contemplated in one way or another.

As HRIM is an information model focused on modularity and interoperability, taking special focus on the components themselves, it's safe to assume a lot of the data we consider necessary is not contemplated by OPC UA, and that we approach this communication differently. However, there are many parallels, including (but not limited to):

  • Their Identification data structure, which serves as a way to identify specific components throughout the whole system, is quite comparable to our ID topic definition, with the same focus. This structure is critical as it applies across-the-board as a way to identify any physical device in the system.

  • Their separation of a Motion Device System (a full robotics system) into motion devices (e.g. robotic arms) and controllers (e.g. a PLC) is comparable to our module classification into composite (specifically arms) and cognition (WIP) modules, respectively. All of the information related to that motion device's state (e.g. whether it's active, or its current pose and trajectory) and identification is already contemplated in some way in HRIM, if not already built-upon an extended.

There's still a lot of work to do; the cognition and communication module definitions, a state machine, safety states and their communication... They are all in the works. However, the structure is there, and even if both information models aren't identical, the vision of both aligns. Our review of OPC UA's current state shows that HRIM is on the right track.