What makes a good project schedule

11 replies [Last post]
Sunchana Johnston
User offline. Last seen 1 week 4 days ago. Offline
Joined: 12 Jul 2009
Posts: 10

Hi All,

I'm just wondering what are the key components for a good project plan / schedule?

I'm thinking the basics 1st, e.g. Start Milestone, Key Deliverables, good WBS, Resources, etc.

Any thoughts? What would you PMO say it's right?

Thanks!

Sunchana

Replies

Rafael Davila
User offline. Last seen 6 days 13 hours ago. Offline
Joined: 1 Mar 2004
Posts: 4384

Network logic as well as many other parameters might change; if predictable a good model shall be responsive to these changes. For many of us it is time to move away from the exclusive use of old CPM models.

1. Conditional scheduling can fix predictable changes in logic.

2. Variable quantity assignments can fix resource demand/supply changes.

3. Automatic resource replacements with other having more/less productivity can fix predictable changes in resource replacement when it makes sense for the activity to start at a higher/lower pace.

4. And the list goes on, it is HUGE !

5. With the assistance of the computer it is possible to create more dynamic schedules.

6. There will never be perfect models but we are setting the bar too low by promoting software and scheduling practices that prevent the scheduler to create better models.

https://www.linkedin.com/pulse/gao-wrong-rafael-davila

https://www.linkedin.com/pulse/usace-cpm-schedule-specifications-rafael-davila

Not surprisingly more technically oriented institutions such as NASA use P6 as a data entry shell and Aurora software as the engine.

https://www.stottlerhenke.com/products/aurora/

7. Perhaps too much to ask for GAO, AACE and USACE schedulers, but not everyone runs that low.

8. It is the end user, the one responsible for the means and methods who shall decide how and what tools he should use. 

Zoltan Palffy
User offline. Last seen 4 days 6 hours ago. Offline
Joined: 13 Jul 2009
Posts: 1472

check your email here 

Rafael Davila
User offline. Last seen 6 days 13 hours ago. Offline
Joined: 1 Mar 2004
Posts: 4384

Ideally there shall only be one agreed to Project Schedule and there should not be different schedules floating around.

This is not always possible, in such case better manage using a Ghost but Real Schedule than an Unrealistic Schedule tailored to satisfy the whims of someone who insists on hijacking the means and methods.

http://www.fplotnick.com/constructioncpm/2016Presentations/MON37-PPR.pdf

Rafael Davila
User offline. Last seen 6 days 13 hours ago. Offline
Joined: 1 Mar 2004
Posts: 4384

A good schedule is not constrained by artificial rules such as the infamous 14 point assesment which has been questioned by many.

http://www.ronwinterconsulting.com/DCMA_14-Point_Assessment.pdf

Rafael Davila
User offline. Last seen 6 days 13 hours ago. Offline
Joined: 1 Mar 2004
Posts: 4384

Using predecessor logic to model crew flow is not a good practice, to the extent many specifications prohibit their use to force resource leveling.

By looking at the links table or by enabling the Gantt to display the temporary links created by the software you shall be able to see the flow of some resources.  For more advanced resource models that provide for intermittent activity scheduling you will need to look at detailed reports for resource assignments, it is not as simple as we wish.

In my previous post the dashed arrows represent the temporary links that the software algorithm used for some of the resource leveling.

Rafael Davila
User offline. Last seen 6 days 13 hours ago. Offline
Joined: 1 Mar 2004
Posts: 4384

Activity duration shall take whatever it needs to be in order to get better schedules. 

Duration of most activities is determined by the volume of work and resource production rates, there is no need to split activities to know progress, it is a lie, do not be lazy and measure volume of work done. 

Artificial splitting of activities is wrong and might lead to undesired intermittent work under resource leveling. Trying to solve the issue using activity priorities will not always work.

Before leveling it looks like equivalent models.

Activity_Splitting_01

After leveling they are not always equivalent.

Activity_Splitting_02

Rafael Davila
User offline. Last seen 6 days 13 hours ago. Offline
Joined: 1 Mar 2004
Posts: 4384

Among the many things that make-up a good schedule is that it can model how a good PM would schedule the activities.

A good Project Manager:

  • Knows about Student Syndrome and Parkinson’s Law.
  • Knows that bloating activity durations is not a good idea as in most instances field managers will absorb all provided activity durations, he knows this is a self-defeating strategy.
  • Knows that in order to increase the probabilities of success he must target for early completion in the hope he will meet contractual milestones after some delays. 
  • Knows about the need to provide for time-risk allowance. 
  • https://gmhplanning.co.uk/wp-content/uploads/2011/10/Newsletter-43-Contractors-float.pdf?8af1b3

Project schedules must consider all existing constraints including resource, supply and financing restrictions. It should model real life work, taking into account working in shifts, activities where execution can/cannot be split, resources that may replace one another, and everything else that is taken into consideration by people that create resource schedules manually. If the schedule model cannot consider real life restrictions, the schedules created by this model will not be practical.

Rafael Davila
User offline. Last seen 6 days 13 hours ago. Offline
Joined: 1 Mar 2004
Posts: 4384

Among the many things that make-up a good schedule is that it can model how a good PM would schedule the activities.

No rookie PM is going to delay an activity because out of 10 resources one will be missing for weeks. The rookie will start the activity at a slower pace and will continue at a higher pace when the remaining resources become available. Using variable quantity the activities will be finished before than if using fixed quantity assignments, in addition idle time will be minimized if using rookie strategy.

In the following example there are 19 resources A & B available. He will make use of all available resources in the most efficient way he can figure out unless the model can give him something better. In complex scenarios a good model will give him a better answer and will provide for variable quantity assignments in a way that makes sense, similar to what he would do manually but more efficiently.

Variable_Quantity_9.2.17.02

Rafael Davila
User offline. Last seen 6 days 13 hours ago. Offline
Joined: 1 Mar 2004
Posts: 4384

Among the many things that make-up a good schedule is that it can model how a good PM would schedule the activities.

In many scenarios materials shall be installed as soon as possible no matter if the activity becomes intermittent.

St._Martin

In some cases it makes no sense to wait until all activity materials are delivered; in desperate situations, and in not so desperate, looting might happens, better install the bricks ASAP. 

In the case of St. Martin, after the hurricane I do not believe an islander would be so foolish as not to install every single brick, piece of wood or many other materials ASAP.  On the other hand some activities such as cement plaster shall be scheduled to be continuous in most instances.

An islander would install bricks ASAP.

Install_ASAP_01

He will never wait for the activity to be continuous.

Install_ASAP_02

Zoltan Palffy
User offline. Last seen 4 days 6 hours ago. Offline
Joined: 13 Jul 2009
Posts: 1472

What makes a Good Schedule?

There are many things to consider when developing or reviewing a CPM schedule. What things should be incorporated into a schedule when developing or reviewing a schedule?

 

This paper will provide a guideline as well as the rationale for the user to follow when creating or reviewing CPM schedules for large scaled construction projects.

 

•          All activities have at least one appropriate predecessor and successor. Only “StartProject” and the “End Project” are                 open ended. The entire CPM schedule must be tied together as one cohesive entity.

 

If an activity does not have a predecessor the Early Start date will be the project start date or the data date whichever is later in time. For projects without a Must Finish By date, activities without successors are assigned a Late Finish date equal to the latest calculated Early Finish date. If the project uses a Must Finish By date then the Late Finish date will be assigned to this date.

In essence if an activity does not have a predecessor or a successor the float values will incorrect. Float values will be too large since the backward pass will erroneously begin at the project completion date.

                    

•           The CPM schedule must model the Design-Access-Material-Manpower constraints inherent to construction projects.

                       

Design - Construction design documents must be timely available to the contract trades for them to develop and submit shop drawings and to put work in place. Excessive RFIs (requests for additional information or clarification) may indicate that the design documents are insufficient. Until the trades can work productively and efficiently (without interruption) - they may resist bringing manpower to the project.

Access  - The trades must have efficient access to the work. If trade workers do not have efficient access to their work, or if materials cannot be efficiently located to the work areas, the trades will tend to avoid working in those areas. In-adequate crane “hook time”, insufficient load/unload areas, excessive trade collocation, improper trade sequencing, insufficient scaffold / man-lift planning can lead to access inefficiency.  Until the trades can work productively and efficiently - they may resist bringing manpower to the project.  

Material - must be timely available. Shop drawings prepared, accepted, materials ordered, fabricated and delivered.  Until the trades can work productively and efficiently (without interruption) they may resist bringing manpower to the project.

Manpower - Sufficient and appropriately skilled manpower must be available. Note however that unless the design, access, and material requirements are met, manpower alone is insufficient.

Efficient progress cannot occur unless all four key elements (Design-Access-Material-Manpower) are adequately addressed. All four are directly related to each other. If one

element is adversely affected, another may also be negatively impacted. It is essential that these four elements are available when creating a CPM schedule.

 

•            The CPM schedule must be able to model subcontractor contract obligations.

Usually the Prime contract obligations transfer a host of items to their subcontractors. Some of these as they relate to the schedule usually are payment, warranties, extras, time requirements for performance, suspension of performance, termination of contract, claims, and dispute resolution. 

The schedule should include the entire scope of the subcontractors as well as the prime contractor and all work should be scheduled to complete within the given contractual completion date by ALL parties.

 

•             Must be able to indicate work flows (crew movement requirements). 

               Crew Flow Activity is a series of scheduling activities that are linked together that apply to a specific area or type of                  construction. Crew Flow logic is applied when these areas exist in several locations within the Project to develop a                   sequencing (or Flow) of the similar areas or types of construction. 

               The use of Crew Flow maximizes resource efficiencies and coordination for all trades. By using a Manpower “ S”                      Curve Analysis for Crew Flow Activities adjustments to activity sequencing can be performed to acquire the ideal                       manpower S curve.

•        Schedule activities should be capable of occurring without interruption from other activities. If the activity will need to               stop so that other tradescan work, then resume; then the activity should further detailed into additional activities to                 prevent this.

 

•         Activities can be measured in terms of units or a commodity i.e. (1000 LF of small conduit, or 400 LF of large cable, or              650 Light Fixtures, etc.) or man-hours.

In order to determine the duration of the activity it is necessary to be able to measure the activity to know some type quantity for that activity. It is necessary to know how much of something there is before a determination can be made as to how long it is going to take to perform that task. 

 

•           Construction activity durations should not exceed 21 working days.

Typically on large projects a reporting period is 1 month which is typically 21 working days. If an activities duration is too long then it would be difficult to estimate the remaining duration when statusing the activity. Using 21 working days as a guideline forces the schedule to be broken down into meaningful, trackable, and measurable activities. This does include long lead time items such as procurement items. 

An exception to this is when developing the schedule using a rolling schedule especially in a Design/Build environment when the finishes have not yet be determined.

 

•           Relationships are Finish-to-Start, or Start-to-Start AND Finish to Finish. SF relationships  are not used.

 

Typically only Finish-to-Start, or Start-to-Start AND Finish to Finish relationships are used in a construction schedule.

 

•            Negative lags (leads) should not be used in a schedule.

There has been a lot of debate over the use of negative lags in a CPM schedule. The use

 of negative lags provides incorrect float vales. There is no such as going backwards in time. Anything that can be modeled with a negative lag can be modeled with a positive lag. Negative lags, or leads, are counter to schedule flow and can make it more difficult to analyze the Critical Path. It is difficult to trace back a schedule path that has a negative lag. This may also indicate that the schedule does not contain a sufficient level of detail.

•             The typical chain for long lead time items and major material deliverables should be identified as separate activities                  in the schedule. 

Activities for Submittals, Review/Approval, and Procurement time that support the installation of the item should be clearly identified so that they may be tracked to ensure that the entire process supports the installation or need by date.  

 

•           Relationship Lags should do not exceed the Predecessor Duration.

Typically the relationship lag should not exceed the duration of the predecessor. When this occurs it results in an unintended non-overlapping activity. For example an activity that has a duration of 10 working days and a successor relationship to another activity with a start-to-start relationship with a 5 day lag. The intent is that half way through the first activity the second successor activity can start. If this lag is greater than 10 days this relationship in essence has changed into a finish-to-start with a lag which was not the original intent of the relationship lag. This often happens when a schedule is created then for whatever reason the person decides to reduce the original duration but does not look at the lag duration to its successor.

•            Weather sensitive work is properly assigned to a “Weather Sensitive Calendar”. Anticipated adverse weather is                         programmed into the CPM network calendar.

Often there is work that must be performed outdoors. This work can be subject to temperature, moisture, freezing and a host of other restrictions. Any work that is weather sensitive should be assigned to a weather sensitive calendar. 

The number of Monthly mean Anticipated Adverse Weather days is typically calculated based off of historical data that can be provided by National Oceanic and Atmospheric Administration (NOAA). This calculation is based on geographical locations and will list the Monthly Anticipated Adverse Weather Delay Work Days. This is often listed in the scheduling specification.

When creating a weather sensitive calendar the number of Monthly Anticipated Adverse Weather days should be indicated as non-working days on a month by month basis. 

 

•           Contract finish milestones are calculated using a Finish On or Before constraint.

In order for the backward pass schedule calculation algorithm to work properly any contractual finish milestones should use the Finish On or Before as its Primary Constraint type.

•           The “As-Planned” CPM network is based on conditions known on bid-day.

When creating the As-Planned Baseline CPM schedule only the information that was known on bid day should be considered when creating the schedule. The As-Planned CPM network represents the contractor’s plan for prosecuting the scope of work known as of bid day. In some cases the project CPM network is developed several weeks after NTP is issued. It is imperative that the CPM network development team build the as-planned CPM network with “blinders” on, so that information received after bid day is not incorporated into the project plan. For example, the contractor may find out after the bid date that the soil on site is contaminated. The additional time for removal of the contaminated soil must not be included into the Baseline As-Planned CPM network. Without a Baseline As-Planned CPM network as a reference point, the time impact for the added scope of work (contaminated soil) cannot be accurately measured.

 

•            Obtain input and gain formal commitment (buy-in) from all project team members (foreman, superintendants, etc).

It is imperative that there is formal buy-in from all parties involved in the project. It is nearly impossible to hold someone accountable for something that they have never seen, reviewed, or agreed to.

 

•              All stakeholders are committed to one active contract CPM network. Separate “Owner” and Subcontractor”.

   There can only be one agreed to Project Schedule and there should not be different schedules floating around. This     will only cause confusion amongst all parties involved.

 

Additional Guidelines

 

Another guideline to adhere to is the Defense Contract Management Agency (DCMA) 14 Point Schedule Metric. This is another way to make sure that the schedule meets certain standards.

The DCMA 14 Point Schedule Metric uses certain guide lines, thresholds, and criteria as standards that every schedule should strive to meet.

These 14 points can be broken down into 2 distinct categories, Quality and Performance.  This paper will briefly discuss the Quality aspect category.

 

DCMA

Point  #     Description                                     Explanation                                         Pass or Failure Threshold

1

Logic

Counts the number of Activities missing a predecessor or missing a successor

Must be Less than 5% of the total number of activities

2

Leads

Check for Negative lags

Must Be Zero

3

Lags

Checks the Total Number of Activities with a lag

Less than 5% of the total # of activities should have lags

4

Relationship

Types

Checks the number and types of FS predecessor relationships to the Total number  of Relationships

At least 90% of the total predecessor relationships must be of the Finish to Start type

5

Hard Constraints

Checks the number of constrains being used in the schedule (does not include Milestones & finish as late as possible is not a hard constraint)

Less than 5% of the total # of activities should have hard constraints

 

DCMA

Point  #     Description                                     Explanation                                         Pass or Failure Threshold

6

High Float Values

This counts the number of Total Activities with high Total Float values

High float values where float is greater than 44 working days. Must be 5% or less of the Total Activities

7

Negative Float

Determines how many activities have negative float values

If any activities have negative Total Float, then this test Fails

8

High Durations

This check looks for activities with too large (or ‘high’) of a duration

5% or less of the Total Activities having less than 44 working days of duration.

9

Invalid Dates

Actual dates should be NOT later than the Data Date

Must be zero to pass test

10

Resources

This test is intended to verify that all tasks with durations of at least one day have dollars or hours assigned

Must Less than 5% of the total # of activities

 

There is also the GAO (General Accounting Office) of the US Schedule Assessment Guide here is the link.

http://www.gao.gov/products/gao-16-89g

In summary there are many different guidelines that can be used to determine the overall quality of a schedule and what makes a good schedule. This paper pointed out several basic items to consider when creating a schedule or when reviewing a schedule.

Mike Testro
User offline. Last seen 3 hours 40 min ago. Offline
Joined: 14 Dec 2005
Posts: 4273

Hi Suchana

It must:

Represent the agreed project sequence.

Use only FS links

No Constraints

Clear critical path

Be flexible to change

Correct calendars

WBS is a side issue for bean counters so ignore that.

Best regards

Mike T.