[TW-1323] Implement Queueing Theory Indicators in burndown reports (Little's Law: TH=WIP/CT)

[taskwarrior]

Benjamin Weber on 2014-05-13T15:39:40Z says:

The burndown report shows me:

 Add rate:         1.1/d      Estimated completion: No convergence
 Done/Delete rate: 0.8/d

 Add rate:         1.1/d      Estimated completion: No convergence
 Done/Delete rate: 0.8/d

 Life Expectancy of a single task: 
 Min: 5 mins / Average: 4.43 days / Median: 13.7 days / Max: 90 days

 Task added every: 6 days
 Task completed every: 7 days 

Note the numbers are placeholders, only. Which would be an excellent indicator of the waiting time of any new task entering the waiting line (which task is).

Think of a pond with water:

!Produc2.gif!

Little's Law (cf. attached pdf for a proof):

!ll.png!

The system has a defined limit, its capacity C.

Swapped variables:

!ll2.png!

"Cycle time is proportional to queue length."

What is more important to show off? A. Throughput of all tasks (how much water gets through) B. Cycle time: time the water stays in the pond C. Work in progress:

What do you want with burndown?

Criticism requested.

Further Readings: http://web.mit.edu/sgraves/www/papers/Little's%20Law-Published.pdf http://iew3.technion.ac.il/serveng/Lectures/Little.pdf George, Michael L: The lean six sigma pocket toolbook http://www.worldcat.org/oclc/56971879 http://courses.cs.washington.edu/courses/cse597/98wi/lectures/lecture5/sld011.htm

[taskwarrior]

Migrated metadata:

Created: 2014-05-13T15:39:40Z
Modified: 2015-03-14T17:49:18Z

[taskwarrior]

Paul Beckingham on 2014-05-13T19:56:37Z says:

This is a discussion of queue theory, but with no mention of it, and several links. I don't know what the request is, or why it benefits the user.

ONE CHANCE: Explain briefly what the request is. Explain the benefit. Or I close it.

[taskwarrior]

Benjamin Weber on 2014-05-13T21:13:18Z says:

If you want me to change the title, enable me to do so, please. If not, no problem - for your convenience:

TITLE: Implement Queueing Theory Indicators in burndown reports (Little's Law: TH=WIP/CT)

BENEFIT: User realises earlier that reduction in tasks WIP decreases CYCLE TIME of a task by maintaining THROUGHPUT rate of tasks. Extension of optimisation thinking from ONE indicator to THREE.

Realisation of:

• THROUGHPUT of tasks decreases with increasing CYCLE TIME of a task.
• CYCLE TIME of a task increases with no. of tasks WIP.
• CYCLE TIME of a task decreases with increase of THROUGHPUT of tasks.
• No. of tasks WIP increases with increasing CYCLE TIME of a task / THROUGHPUT of tasks.

CYCLE TIME: Time to complete a task THROUGHPUT: No. of tasks for time interval

Is the title better now? I appreciate your kind input to lift any fog.

Disclaimer: If you really believe, I do this for fun, feel free to delete my account.

[taskwarrior]

Paul Beckingham on 2014-05-13T22:26:07Z says:

Great. Thank you.

[pbeckingham]

TW-1323_Little.pdf.txt