Project Planning

Discussion in 'The Projects Forum' started by Insertname, Sep 8, 2011.

  1. Insertname

    Thread Starter New Member

    Mar 15, 2011
    12
    0
    Don't shun me! Hear me about before you just look down on this as 'management <snip>' or anything like that. I'd like serious helpful replies here from people who have experience planning, scheduling, and sticking to schedule with their projects. I know projects are complex and have many variables but planning can and does help (it helps me, but I'm not great at it) but I need to learn more on it. So rather than googling the dreadfully boring theories, I decided to ask people who might actually apply it.


    I'm open to all advice, but specifically I'm interested in the following:

    1. I know you can schedule tasks that have well defined times associated with them, but how do you plan for and schedule a project you wish to complete that you know you'll need to learn a lot for? Since you may have no idea on some parts of the projects, how do you even begin to schedule those in any slightly accurate way?

    2. How long should planning the project take? I know this is a bit of a 'who watches the watchers?' question, but I thought it'd be worth asking.


    Thanks in advance all! :) Any help much appreciated
     
    Last edited by a moderator: Sep 12, 2011
  2. wayneh

    Expert

    Sep 9, 2010
    12,087
    3,026
    The traditional way is to go find somebody that does know the work, and enlist their help. A schedule produced by a newcomer (typically a manager who doesn't really know or respect the work) is close to worthless. I've seen many of them, and they are uniformly crazy optimistic. Wishing a project to completion by writing it down rarely works.

    You might also find a forum more tailored to this topic, this is an odd thing to post on a forum devoted to circuits.

    One last word of advice:
    Give me six hours to chop down a tree and I will spend the first four sharpening the axe. Abraham Lincoln
     
  3. MMcLaren

    Well-Known Member

    Feb 14, 2010
    759
    116
    Business school classes called Operations Management usually cover alot of ground, including Gantt charts like the example below.

    Regards, Mike

    [​IMG]
     
  4. wayneh

    Expert

    Sep 9, 2010
    12,087
    3,026
    Yup, that's one widely used tool, along with PERT charts and the like. But the age-old database axiom still applies; Shinola in, shinola out.

    One thing I've found critical is to assign the tasks on the chart to some individual, and get agreement from that person that the proposed schedule is acceptable. So instead of just recognizing the time a task might take hypothetically, the proposed schedule actually reflects current conditions that might extend the task. Timelines merely imposed from above without input get the respect they deserve - not much.
     
  5. Insertname

    Thread Starter New Member

    Mar 15, 2011
    12
    0
    I posted this in a circuits forum because a lot of people here have undertaken their own projects before so would be more in the know when it comes to forecasting how long a project will take, even without getting in to WBS, critical paths, and the like (I had to study those in 3rd year of college, but none of it was at all meaningful).

    Thanks for the input on getting specialists involved for the project. I always suspected just having a generic 'manager' wouldn't cut it when it came to making working plans.
     
  6. wayneh

    Expert

    Sep 9, 2010
    12,087
    3,026
    Dude, that says far more about you than about the subject. ;)
     
  7. MrChips

    Moderator

    Oct 2, 2009
    12,414
    3,353
    Rule of thumb - take your best estimate of the time required and then double it.
     
  8. Dr_Ram

    New Member

    Sep 14, 2011
    7
    2
    I will not attempt to answer your question in complete generality, but after spending 17 years in the IT industry and an even longer time in planning my DIY electronics projects, if there is only one aspect of planning and scheduling my project that I would choose to remember, it is this (may not work for you but it works for me - and it worked even when I was preparing for my graduate work in General Relativity which I was taught by a student of Stephen Hawking): Identify the 2-3 most difficult/fearsome/unknown/unpredictable aspects of your project, plan,schedule and attack them first. For my electronics projects it has always been the ability to house it in a quality chassis with a professional front panel. Therefore I would spend ALL my energy solving that problem first, after envisioning what front panel my project would need from a "user point of view", THEN I would get down to designing the actual circuit.

    Cheers - Ram
     
  9. DickCappels

    Moderator

    Aug 21, 2008
    2,647
    632
    Answer: You can't estimate accurately. Make the best estimate you can based on what is known and using the best advice you can get, and then make provisions for reviewing and resetting or stopping the project if it becomes too much. You can't know more than you can know (just finished a quote that pretty much says the same thing).
     
  10. russpatterson

    Member

    Feb 1, 2010
    351
    16
    Break it into parts that are 2-4 hour chunks. Since you can't know ahead of time how long it will take to learn something that is unknown to you, you add variability. E.g. this task is from 2-8 hours. Then you get a high and low range for your estimate. Then you re-do the estimation process half way through and you'll have a pretty good guess. Know what you don't know.
     
  11. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    You'll never know (as DickCappels said).

    Even with years of experience you get one job that should have been easy but many unforseen things cropped up and totally blew out the timeline. Then sometimes you get a job that should have been hard but for some reason it just all falls into place quickly and easily.

    Having said that it just about always takes longer than you first would have estimated, so MrChips' suggestion of giving it your best shot then doubling it is not too far off.

    Or maybe give it your best estimate, then roll a dice and multiply your time estimate by the number on the dice! ;)
     
Loading...