Distributed Orchestration
The Paranet supports centralized, decentralized, or distributed coordination, where actors (e.g., AMRs, humanoids, agents) self-orchestrate or delegate tasks locally, minimizing latency by deciding closest to the knowledge source (robots) or best source of knowledge (agent). All paranets will have orchestration actors. They represent the intention of the actors in their nodes.
By analogy, using a collaboration application like Slack, there is no overall controlling orchestrator. Users send and respond to messages. Slack is a kind of paranet where every user is an actor making decisions on their own such as what to send, what to respond to, etc. Slack is a ledger of messages. Each channel is like a node. The Paranet is a generic kind of collaboration system/network where actors can be any kind of intelligent entity with a digital interface.
Every organization has heterogenous types of assets that can be orchestrated more effectively using the Paranet. This is because they can all work securely on a network where the assets provide their intelligence in the form of skills they have and need from others. When these assets become actors (e.g., python code running robot) by running on the network, their skills are discoverable to the network. An actor can say it needs another actor to perform work and the network makes this happen by engaging the two actors in a written (persistent) collaboration. This is how Slack works. Someone writes a message, and others respond though Slack does not have formal orchestrators—there is no plan or purpose behind Slack. But on the Paranet the messages are the formal semantics of the organization where actors are performing work, not just writing messages. And any asset can be an actor (ERP, ticketing system, forklift drivers, and so on).
Given that paraflow actors are planners, and that they run on a network and delegate work, plans can span an unlimited number of tasks to be accomplished by an organization. Every actor has skills and knows what it is supposed to do, but none can work alone and be truly autonomous, each needs to be part of an organized, orchestrated network.