Something inbetween Work Order and Project?

[fosterlynn]

Do we need something between a Work Order and Project as an Agent?

Sensorica uses Project as something fairly ongoing, with some of its own governance (like value equation), roughly relating to a product line. It has a strong organizational flavor, and we have slotted it as an Agent.

Many people use the word "project" as something that produces a goal, has an end, and is not primarily organizational. This is common in IT for example. And it came up when discussion how to structure the NextEdge working group data. This however might be larger and longer than a work order, and work orders are needed for smaller units of work.

Should we add something else? If so, how to model and how to represent in the UI?

[bhaugen]

Steve Bosserman was also talking about something like this, I think, for the PV characterization project, which has a bunch of related work orders, but you can't see the relationships in NRP.

[bhaugen]

Lynn and I tried to list some spectrum of stuff that people call "project", starting with the smallest, noting what this is in NRP (if it exists there):

• A single process that more than one person is working on, with a single deliverable. When the deliverable is finished, so is the process.
• Work order: a sequence of processes leading to one or more deliverables. When all the deliverables are finished, so is the work order.
• Something between work order and a Sensorica project, which does not exist in NRP, and which is the focus of this issue.:
  • for example, Tammy has projects that involve several people creating some media assets.
  • Steve's idea seems to fit in this middle ground, too: a set of work orders within a Sensorica project that are all related.
  • or, possibly larger, some open source software projects. They have a deliverable (their code), but the code goes through revisions and different releases and branches and adds features and potentially goes on for years. The project has different levels of membership: for example, committers, forkers and other contributors. Might have sets of rules.
  • more examples in this middle ground would be useful.
• A Sensorica project: has a focus, and a group of members. May have many work orders with many deliverables. Might have its own value equation and other governance features. Might have sub-projects, or might be a sub to another project.

One of the distinctions in NRP is whether such a thing is an Agent or not: that is, does it have agency beyond just a group of people working together to do something? Signs of agency would be: having its own value equation or other governance rules. Or, doing exchanges with other agents.

If so, it might also be a "context agent", which is one of the concepts in NRP that seems to confuse people, and for which we don't have a coherent explanation yet. But if so, then processes and work orders may happen in the context of this project.

So what else does project mean, and where does it fit in this spectrum?

[stevebosserman]

Perhaps a way to look at this would be to designate programs for more macro initiatives within which there can be multiple projects each comprised of process-related work orders. Both industry and academe oftentimes subscribe to "projects within a program" in organizing their work. And there are distinct leadership qualities and management skills associated with each. Just because one can handle projects doesn't necessarily mean that one can do similarly with programs. As a result, the range of skills required to successfully address projects or programs could be significantly intensified within distributed, collaborative and transparent contexts.

[fosterlynn]

Car and walk discussions:

• We think we do need this concept in the NRP. (Although it's hard to figure out a name. Hey, maybe "initiative" from @stevebosserman comment would work?) Admittedly, this adds yet another complexity to NRP, but this one keeps coming up. And we don't want to force people who do little-p projects to use agents for this.
• It could group both processes and exchanges.
• We could create a type object so people can define their own. Like Crowdfunding Campaign, Project. This could be what mostly shows up in the UI as naming.
• We could add this filter in various places.
• Different networks will have varying definitions of what this is, but that is fine. It is part of modeling your organization, and could support anything between work orders (which can be pretty small) and agents (which have specific behaviors allowed). The project/campaign would not support agency behavior, but otherwise would be pretty flexible in terms of length of time, end goals or not, etc.

[TiberiusB]

SENSORICA uses "project" as a context of work, as defined by Bob at the beginning. I think is a good definition and it sticks with us. A project in SENSORICA, as Lynn mentioned above, is an activity that has a beginning and an end, which is usually a market exchange, or some inflow from outside, like funding (grant, sponsorship, ...). I guess the most generic criteria is whenever the network has an intake as a result of an activity, a sale or some sort of funding, we need a project. Some funding activity can already be part of an existing project, in that case we simply create a process or a Workflow recipe for that same project. SENSORICA uses "Building SENSORICA" as a context of work, or project, for everything related to development or maintenance of the network. For example, if we have a fundraising activity for SENSORICA, we create a process or a Workflow recipe for the "Building SENSORICA" project. From my experience, I think processes do the in-between job...

[bhaugen]

Let's say a Sensorica project has a fundraising campaign, and has a bunch of work orders that corral processes that contribute to the fundraising campaign. That is an example of some set of processes that are in between project and work order.

(But @TiberiusB it's good that you going thru the issues and adding more comments...)