3d visualization for Project Management

Proposal and Request for Comments:

A method for (SL) 3d visualization to be used in (RL) project management
-------------------------------------------------------------------------------------

Background:
------------
The project mangement body of knowledge (PMBOK) guide identifies 9 major considerations for managing projects effectively:
* What are we going to do? (scope)
* Who are we doing it for? (communication)
* How long is it going to take? (time)
* What do we have to buy to make it work? (procurement)
* How much is it going to cost? (cost)
* Who's going to do it? (HR)
* What can go wrong? (risk)
* How do we make sure it doesn't go wrong? (quality)
* How do we bring it all together? (integration)


Many project managers focus on the project schedule. We build gantt charts, but not communication plans, project charters, risk management plans, etc.

Gannt charts are very effective at conveying schedule and work effort, but not effective at conveying risk, cost, resource allocation, etc.

I propose a method for 3d visualization of project plans which better conveys risk, cost, resource allocation and which facilitates communication.

By placing "virtual blocks" with multiple visible attributes (size, shape, colour, spin, location, transparency) into a shared virtual space, we may convey more information about cost, risk, resource allocation, etc, than with a simple gantt chart. Additionally, the nature of the shared space facilitates communication and collaboration.


How is it done?
----------------
Project Manager meets with project team inworld. Each team member identifies work packages that are required in order to complete the project. One virtual block is created to represent each work package. The block is coloured, sized, and placed according to the preapproved methodology (see Block Attributes)

In general, blocks are placed in a starting pile at the start line, then moved along the floor toward the finish line as work is completed. Swimlanes identify ownership or responsibility.

The project manager and stakeholders can enter the shared space and immediately see how well the project is progressing.


Block Attributes:
-----------------
Size: Work effort

colour and transparency: schedule and prioritization
semi-transparent: future work
green: on track
yellow: urgent
red: extremely urgent (holding up the project)

spin: risk or confidence (note, this requires a script to set the spin speed)
- rapidly spinning blocks are very high risk and should be given special attention

location:
X=completeness
Y=ownership (cf: swimlane diagrams)
Z=dependency (blocks perched on top of other blocks - top block must be moved/completed first.)

Shape: type of task (from workflow or business process diagramming standards)
- cube - standard work package
- diamond - milestone
- sphere - ???
- cone - ???
- torus - ???



Benefits:
---------
Virtual teams can meet together in realtime and discuss work packages directly.

Piles of objects representing work are more intuitive, understandable and compelling than more abstract representations such as gantt charts.
(It's cognitively immediate! - it's obvious that a big pile of blocks represents a lot of work.)

Disadvantages:
--------------
Piles of blocks don't convey dependencies very well.
- Stacked blocks can be set up such that the top blocks must be moved first (perhaps toruses on a tower of hanoi). In this case, blocks on the bottom are dependent on those ontop.
- Alternatively, any given block could serve as a platform for a whole series of other blocks (a sub project). The recursive subproject model only supports one or two levels
- Finally, SL "hinges" could be used like wires linking multiple blocks, thus illustrating dependencies.


Work to be done:
----------------
Run through pilot for two or three simple projects to refine methodology. (colour, shape, size, etc)

Develop a method of importing MS Project schedules into a virtual world, to be represented as a pile of blocks.

Develop a method of converting simple lists into piles of blocks. Lists on notecards are very quick and dirty ways to begin developing work breakdown structures.
- Can LSL parse XML from notecards, or must it be Comma seperated values?

Develop a method of sharing documents with the project team (virtual filing cabinets?)

Develop a method of brainstorming together (shared whiteboard) for problem solving, to facilitate understanding and convey meaning.

Improve the messaging and workflow.
- develop a way to physically hand something to someone.. visualize the handoff. This represents workflow.
- develop a way to capture approvals.
- improve the IM interface by adding email with folders?
- develop a way to communicate via voice. Possibly through API with skype or gtalk. (someone suggested a conference server for a shared space (room), then everyone talks to the room.


Improve the calendaring aspect of virtual worlds.
- Scheduling Work: Work packages should be able to message the owner and the project manager when they become due or overdue.
- Scheduling meetings: Groups should be able to host private "events". Events should message the required attendees.

Resolve the dependency visualization problem.


Feedback:
---------
???

Sounds like an interesting goal, but if you haven't read Tufte you ought as he knows quite a bit about visualizing information. More importantly, he has some very good explanations about stuff we'd like to visualize but cannot find a "language" for - which I fear your project may fall into.

Fantastic! Thanks for pointing out Tufte. I haven't read his work, but I'll get it immediately.

There's also an element of collaboration and process to consider.
The team has to be able to build the visual display together relatively easily.

Here I would point to No More Teams by Michael Schrage, even though I haven't read it yet .

Serious Play also sounds pretty good.

3D Gantt???? hmmm

http://www.dundas.com/Products/Chart/Features/3DGanttpopup.htm

Does anyone know if Michael Schrage has done anything with/in SL?

It seems to be the ideal environment for rapid prototyping like he advocates.

Agree with you there. Analog knowledge representation certainly has those practical advantages. But it's sometimes hard to design an analog rep'n that maps to First Order Logic. As you're probably aware, that matters because FOL is precisely understood. If you have a formal specification of your problem, you can know whether or not a given result is a solution. (soundness and completeness)



When you're deciding how to represent dependencies one simple guideline is to make sure that your system can represent the logical OR relation. This takes care of the case in which either one or the other or both of the subtasks must be completed first but we are not able to specify any more details about which one it will turn out to be. The exclusive OR is much simpler and easier for an analog rep'n, but the logical OR can be an interesting challenge.

Have fun!

I wasn't aware of this. I don't have a solid grounding in First Order Logic. I approach this from a business perspective. One of the tenets of project management is that problems can by clearly defined. That may be why traditional project management fails for those "wicked" problems (situations where the problem isn't defined and the process of finding a solution affects the definition of the problem). Most project managers will allow for risk. That helps with those projects where the problem isn't clearly defined and the solution isn't clearly understood.



That's a remarkable insight. I can see how that's vital in computer science, logic and graph systems. I've never seen a project manager map a logical OR into their plan. I can see where it might happen, but they probably represent them as a single work package ("find a solution to X" could mean this OR that).

I'm impressed by your breadth of expertise, clearly different from mine. I'd love to talk about this in-world sometime. What's your background?

-WW

Professional swamp monster. People pay me to get away from their camping chairs. : )

And I know a bit of artificial intelligence, but it's only the academic sort, not the kind that powers the enemy hordes in online combat games.