Planning a Project

1
Planning a Project

• Dr Simon L Kendal

2
Why Plan

• Projects are long, unscheduled, uncontrolled
• Unless you control them
• .YOU MAY FAIL or certainly not do very well
• .you will run out of time at the end and throw
  away lots of marks.
• not enough time to evaluate success of project
• not enough time to write up report properly

3
Why Plan

• We assess control and these are the easiest marks
  to get!
• 2/3 rds of students get poor marks in reviews!

4
Heading For a Fail

• Typical 1st review session
• I have finished my research I have read ten
  books!
• 2nd review
• I am a bit behind but I will catch up
• i.e. havent finished research, not written
  research chapters, still not documented user
  requirements (so I dont really know what I have
  to develop), development of system started but
  not much done yet, still not planned dissertation
  (so I dont know what I am writing) - but I am
  still expecting to finish on time.

5
What is Typical

• A last minute rush to..
• finish the system (result a few bugs left in)
• finish the report (result some sections missing -
  no time for feedback)
• get sponsor evaluation (result not done and marks
  thrown away)

6
Good, Bad or Plausible Schedule?
7
A Common Mistake

• Immediately generating a (plausible) schedule!
• Guessing the dates will not do - you need to know
  they are accurate, calculable, fixed or you will
  not be concerned when you miss them. It was only
  a guess anyway!
• How do we get those dates?

8
How Long Does it Take To

• In real world we need to ask How long does it
  take to
• In projects we can ask How long can I afford to
  spend doing...

9
Steps to Producing a Good Schedule

• Break down tasks on TOR
• Allocate hours
• Plan personal diary
• Work out necessary sequences for tasks
• Allow for known delays
• Allow for elapsed time
• Produce Gantt chart
• Produce Schedule

10
Breaking Down the Tasks

• Look at TOR break each major objective into ALL
  of the tasks needed to be done to satisfy that
  objective.
• Example (note this is NOT an exhaustive list)

11
Assuming a TOR is as follows

• Research topic A
• Investigate topic X
• Develop prototype system to..
• Evaluate project
• Write dissertation (sometimes unstated)

12
Research topic A

• Read introductory material
• Lit search
• Read papers
• Write reports
• also ???
• Interviews?
• Questionnaires?
• Write to companies? etc

13
Investigate topic X

• Read books
• Write example program code
• Interview relevant specialists

14
Develop prototype system to

• Obtain user requirements
• Design (inc. throwaway prototypes)
• Evaluate prototypes
• Redesign
• Develop
• Test / Debug
• Write manuals
• Install

15
Evaluate Project

• Decide on evaluation criteria
• Plan feedback process
• Collect feedback from sponsors
• Undertake user trails
• ???
• ???

16
Write dissertation

• Plan dissertation (i.e. contents page)
• early enough to get feedback
• Write system design chapter
• Write system development chapter
• Write draft dissertation (allow time for
  feedback)
• Write final dissertation (allow time for
  printing/binding)

17
Allocating the Number of Hours

• Assuming 225 hours split among project objectives
  e.g.
• Topic Hours
• Research topic A 75
• Investigate topic X 20
• Develop system 75
• Evaluate project 15
• Write dissertation 40

18
Allocate hours to tasks

• e.g. Research 75 hours
• Task Hours
• Read introductory material 20 10
• Lit search 15 10
• Read papers 40 40
• Write reports 20 15

19
Planning Your Diary

• Allowing for busy and quite weeks plan number of
  hours spent on project

20
Network Charts

• Allow for parallel tasks, gaps and elapsed time
  e.g. research
• 1) Read introductory material
• 2) Lit search
• 3) Read papers
• 4) Write reports
• Elapsed time e.g. five 1/2 hour interviews - may
  take all week and reading papers may take a month
  or two.

21
Producing a Gantt Chart

• Add tasks to Gantt chart - allow for elapsed
  time, gaps and total hours per week e.g. Reading
  papers

Tasks 5 4 3 2 1
10
10
10
15
15
Total 225 hrs
...
0
22
Finally the Schedule

• The Gantt chart will finally give you planned
  start and end dates
• Having taken all factors into account, inc. your
  personal diary, you should be able to rely on
  these dates.
• Transfer this data onto your schedule.
• Stick to these dates!
• even if you decide to spend extra time on the
  project each week.

23
Some Final Thoughts

• Time per week does not really matter, within
  human and physical constraints, but meeting those
  dates DOES matter.
• Allow a short time at the end of the project for
  last minute hitches.
• Reschedule when major changes occur.
• This is your project - not the sponsors - stay in
  control.

24
Conclusions

• Go through each of the steps given to produce a
  good quality schedule.
• Project planning is a tool to help you - dont
  spend 200 hrs planning the project and 25 hrs
  doing the project.
• Monitor the project - Having developed the Gantt
  chart and schedule - USE THEM!
• Control and succeed - Best wishes!