The required reading for this week is Article 7, the Great Divorce: Cost Loaded Schedule Updating, by John P. Orr. This article starts with explaining the impact of resources on schedule, then goes on to resource related problems in scheduling. It is a little long (15 pages including a title page and a TOC) but it ties up nicely a number of things we have discussed. For a simpler explanation of resource constraints on project schedule, read Article 1.
Article 4 is another good article on resource loading and leveling. It starts with a discussion of float, which is our topic last week, then covers the processes of resource leveling. The author Michael R. Nosbisch, is the current AACE President. His co-author, Ron Winter wrote our required reading last week.
Article 5 deals with some practical issues with resource loading in P3 and MS Project. Article 6 is an academic discussion on resource constraint. You may find it difficult to read starting from page 4. Proceed with caution. Article 2 and 3 are on the specific issues of labor resource management. I would recommend Article 3 over 2.
By now you probably have noticed that a lot of these “better” articles are from a relatively small group. Just take a look at the familiar names below, Just, Chen, Winter, etc. Sometimes you can actually follow an individual’s work. Or better yet, you could become one of those individuals/leaders if you are willing to share what you have learned in your profession. You don’t really need an advanced degree for this.
- The Effect of Resource Constraints on Project Schedules, Michael R. Just, CCE; James P. Murphy, AACE Transactions, 1994
- Labor Resource Planning for Construction Projects, Anubhav Bhargava; Balkiz Oztemir; Allan D. Chasey, AACE International Transactions, 2002
- Applying Trapezoidal Technique in Resource Management, Mark T. Chen, AACE International Transactions, 2006
- Managing Resource Leveling, Michael R. Nosbisch, CCC PSP; Ronald M. Winter, PSP, Cost Engineering, Vol. 48, No. 07, 2006
- Artificial Resource Loading for Schedule Review, Albert Joseph Kuhn, PSP, 2007 AACE International Transactions, PS.17, 2007
- Advanced Scheduling Technology for Shorter Resource-Constrained Project Durations, Robert Richards, AACE International Transactions, 2008
- The Great Divorce: Cost Loaded Schedule Updating, John P. Orr, PSP, 2011 AACE International Transactions, OWN.718, 2011
The Effect of Resource Constraints on Project Schedules by Michael R. Just, CCE and James P. Murphy discusses the effects on a schedule when resources are not used versus used and the big difference this can have on the outcome. When project schedules fail to consider resource constraints critical path concepts can possibly break down. Compared to 1994, I would assume many more projects use resource constraints today just for this reason?
ReplyDeleteResource constraints consist of site constraints, competition for resources (labor: maybe not now but think building boom), and geographical constraints. The authors indicate these can be handled by overtime, additional shifts, or subcontracting. Just because the contractor has ample labor resources if the material and equipment are not at the site crews will not be able to perform and a labor profile with peaks and valleys may ensue versus of a more leveled one, hoped for. This creates more headaches calling subcontractors back which I have heard from GCs is a pain and increases costs.
The authors define float as the amount of time that an activity can be delayed without impacting succeeding activities (free float) or the project completion (total float) (DCL.2.1). They discuss and show in the figures that float can be used to shift (I’m assuming non-critical?) activities to adjust resources. If the entire float has been used and there are still insufficient resources than delays occur. The Primavera software examples shown in some of the figures and labor profile highlight what happens when resource leveling occurs (shifting activities around by using float) that demonstrates quite clearly how significant the changes are when resource constraints are used. I wonder if a contractor did have unlimited workers would it still be wise to stretch the project out some more versus the hypothetical earliest completion because cost might be higher and if contractors are on top of each other may create some un-pleasantries at the job site or possible have unforeseen issues crop up?
In the examples, they demonstrate in two instances how using resource constraint, this case labor, the activities that once had float now appear critical and project completion date has been bumped back by “x” number of days. It is interesting to see how an individual can create two totally different schedules by incorporating labor or not as a constraint. I am reminded of an earlier discussion this semester how a recent graduate who is less experienced can make nice looking schedules that are ultimately useless if key items are not included. The schedules in this article are in black and white and I am assuming the hatched or solid filled bars indicate critical activities unlike more recent articles that are in color, are easier to read (decipher for a layman), and I guess shows the natural progression of the software over 17 years.
Near the end of the article is a brief discussion about the contractor on public construction projects preparing ‘early’ schedule that indicates a completion date earlier than the required contract (DCL.2.4). It is indicated this is done in order for the contractor to lay the ground work for delay claims if the work is done on time or early. Does this seem to be bias? Why should we think the contractor is trying to make a quick buck, especially for a project that if using resource constraints must be a good size project and one that requires a good track record by the GC?
It is suggested that the owner’s team review the schedule to see if it is realistic and resource constraints are used or considered before accepting it as the baseline schedule. It proceeds to state, if actual duration exceeds planned duration this may be attributable to the contractor due to poor project planning if resources were ignored (DCL.2.6).
I guess it seems logical to always consider resource constraints on public, federal, and military projects as well as any project over $1 million (or should it be less) so having issues that crop up without thinking about resources does not occur?
DS
Part 1:
ReplyDeleteThe Great Divorce: Cost Loaded Schedule Updating by John P. Orr, PSP based on my understanding indicates that with the advancement of project management /scheduling software that a greater need to comprehend how to use the features effectively, such as breaking apart the time (resource) and cost is necessary because the cost-loaded schedule may not be accurate when determining progress payments. The author mentions the author Murray Woolf from last week’s class that expresses his concern that these methodologies may end up competing with each other on the schedule.
The section Purpose for the Project Execution Schedule is a well-reasoned look at why we need scheduling software in the industry so the contractor is not scratching his head at what to do on a daily basis and have the time to purchase items with long lead times. Resource loading is discussed here and why this is important such as so not exceeding manpower per day, being able to level this and see the ramifications of it over the course of the schedule and the possible effect on the completion date.
The section Schedule Projections Adjusted by Resource Leveling shows histograms of the leveling that occurs when the 20 person team manpower is exceeded on a daily basis per concurrent activities and the effect it has the schedule and most likely cost. Figures 1 and 2 are helpful and the schedule instead of the histogram is more interesting. I am assuming the non-critical activities have been leveled so it does not affect the critical path. Even though the author mentions there has been no imposed finish date which I assume it then makes the CPM less important or non-existence? Not being familiar with Primavera P3/P6 it is interesting to see the scheduler has control over the resource leveling and can click on items in the software to enable it or disenable it. It is a reminder to me that the individual controls the schedule and not vice versa (as long as they are intelligent and understand what is going on to accurately create the schedule.)
The section Cost Loading as the Basis for Progress Payments dives into divorcing the time and money when a project is in full swing and for example all parties agree 90% is done out of a 10 day activity duration even though all parties are aware more than 1 day of the activity still exists. Being able to unclick the schedule percentage and cost percentage makes sense and is good to know when and if I encounter this software that this exists.
The example on a section of damaged pipe or a missing hanger that really is not holding up the project but because the owner’s project team is unwilling to sign off on it at 100% makes sense but can and will throw off schedule projections is good to know. The discussion that ensues about ‘out-of –sequence’ reporting and issues that arise about not having 100% but 95%-97% makes successor items finish later when they may not need to do. This may affect record-keeping and any time impact delays evaluation.
The Situation: Distorted Schedule Projections section elaborates on methods of dealing with out-of- sequence reporting. Although I find this a little difficult to follow because I have never used scheduling software in an office, and have never needed to focus my attention on this, I think it implies this: Primavera allows retained logic, progress override, and actual dates that maintain the CPM calculation until out of sequence occurs. Retained logic allows successor items to start but uncompleted portions maintain a ‘wait’ until all predecessor work is completed. Figure 3 shows necking (narrowing bars), indicating non-work periods while the predecessor items are completed which I think means activity finish projections are skewed and superintendent and contractors cannot rely on them.
Part 2:
ReplyDeleteThe distortion can be corrected by using progress override or actual dates but in many cases are progress override is prohibited in contracts. This is prohibited because original schedule logic is disconnected from all out of sequence events without regard to if it is still valid and impacts the CPM.
The Case Study- Shift between Retained Logic and Progress Override, I think makes sense. It seems to be that the contractor eventually opted of the progress override after the mid-way point and relied too heavily on concurrent activities when crews were to thin and the owner kept his own shadow schedule that relied on retained logic. In both cases the schedule did not provide an accurate projection of the project completion.
Recommendation: Divorce the Cost Factor, indicates that focusing too much on cost progress or schedule progress disrupts the other one. This is why a divorce is needed. The author recommends focusing on schedule progress (time aspect) because if this falls behind than all other aspects are no longer valid. If a activity is not necessarily completed but far enough along that all other successors can start, than treat it as 100% and keep a running punch list which will cause less disruptions of the remaining scheduling duration. I think this makes sense and seems to be more sensible than what was previously discusses, particularly in the case study. Figures 4 and 5 graphically show this and are good in seeing the two items that have had payments withheld down towards the bottom of the schedule. This ties into figure 6 Earned Value Report and shows the discrepancy between schedule percentage completion, 65% based on cumulative earned value and not actual cost which would be 65% cost variance.
In conclusion it seems like it is a delicate balance, almost a science in understanding and figuring out what is the best approach. Being an individual who has never needed to do this part of the project, I find it daunting at times. I have to jog my memory on our discussion of Earned value method from a couple of weeks ago for the last part of his discussion, where if these were terms I used on a regular basis, most likely it would be clear as day. Who in an office decides if this divorce method is going to occur on all projects or on some and if many recent graduates are doing scheduling should the office expect them to know the ins and outs of such matters when it seems this is geared more for an experienced professional decision maker? The last tidbit of advice is that there are only two times a schedule progress percentage update must match the cost update: at the baseline (zero progress) and at the as-built (100% progress) (OWN.718.15)
DS
CMGT 6630
ReplyDeleteBlog for week 11
Steve Nellis
Required reading: The Great Divorce: Cost Loaded Schedule Updating
This article presents the author’s suggestions that the “divorce” of cost tracking for schedules is the preferred methodology based on the problems inherent in using cost-loaded schedule and the resulting distortions that can occur.
The article starts by discussing the concept of resource allocations in the scheduling process. It states that “forward leveling determines the earliest possible dates for scheduling without exceeding the maximum amount of resource availability.” The illustrations that are provide help to clarify his statements. In high-production home building (HPHB), this is a concept that is not discussed or used. This is because each house is on a very short cycle time and the resources that are necessary are left to each of the trade contractors to determine and manage. Now I can see how large trade contractor organizations will utilize the concepts are provided, especially since they serve multiple home builders. However, due to the constant fluidness of each home builder’s superintendents needs for resources, this type of scheduling is not very practical.
Later in the article, the author discusses that an inherent and frequent problem in updating cost-loaded schedules is the decision to withhold payment just under 100%. In our world of high-production home building, we usually pay trade contractors in draws. For example, the plumber will get three draws – one for rough-in; one for top out; and one for trim out. The total of these three draws equals 100% of the total contract payment. However, in no case will we knowingly and willfully pay one of these draws unless that draw portion of the work is 100% complete and accepted.
Later in the same paragraph, the author talks about keeping a percentage of the contract amount unpaid prevents the defective or missing or temporary accessory work from being forgotten. However, in this scenario, the burden of responsibility now shifts from the contractor to the superintendent / project manager.
In the next paragraph, the article states that for an activity that has no further impact on a project’s schedule could be marked 100% complete. Again, I disagree because it will have an effect on the schedule later on if the remaining work is not completed. Also, for the publicly traded HPHB, there are internal audit requirements that have to be adhered to. During these audits, a team of auditors will visit job sites to compare them to the reported schedule. Since our schedules do in fact pay contractors immediately upon the “click-off” of the event, any “Click-offs” that are not supported by the field visit are extremely scrutinized and the responsible party is subject to disciplinary actions, up to and including termination. In fact, I have had to terminate an employee for this very reason.
The author then presents two recommendations; the first one I am in agreement with, the second one I am not in agreement with.
The first recommendation states that when the cost-loaded schedule update is used to generate progress payments, the attention of the project team is unavoidably directed toward its purpose as a cost reporting tool. This is an accurate statement in HPHB. It is used as a cost reporting tool. However, the author continues to recommend that if the project management team focuses too much to cost progress, the update inputs can distort the schedule’s accuracy. I believe this to be fairly accurate statement. However, with such short cycle times in home building (30 to 90 work days), this is hardly an important factor.
CMGT 6630
ReplyDeleteBlog for week 11
Steve Nellis
PART TWO
The second recommendation is to treat an activity as 100% complete when it has reached the point where its successors can begin. This is not acceptable behavior in HPHB. The best scenario I can present will illustrate this. In the case of a home, we are legally substantially complete once we obtain a Certificate of Occupancy from the governing agency. Technically, there are not any additional items that are scheduled. However, in the eyes of the customer, this house is not 100% complete until we have completed the pre-closing orientation and all of the punch list items have been completed to their 100% satisfaction.
In the conclusion of the article, the author states his position for the divorce of cost-loaded scheduling. Since I am not in agreement with his ideas, I do have an alternative that is widely used in construction – the concept of retention. In the project schedule, if each activity were to have a second sub-activity for final retention payment, this will alleviate the concerns that I have highlighted while accomplishing the concepts that the author presents.
Second Article Reading: The Effect of Resource Constraints on Project Schedules
This article was rather short in demonstrating its pretty apparent and obvious point – that proper resource management is important to the success of a project. They provide an example of a project and illustrate different scenarios of improper versus proper resource management for the project. All of this makes sense, especially in the world of large, single commercial or industrial projects. Again, not much applicability to HPHB since our cycle times are so short for building individual houses.
The conclusion of the article pretty much sums up the ideas that are discussed. The differences in resource constrained schedules and schedules that do not have any resource constraints can be significant. Unanticipated overtime adds costs, schedule delays will occur, and the resulting overruns are all results of planning and scheduling without the proper consideration of the available resources. If resource considerations are not used during the project’s planning, they should always be considered when reviewing the history of the project when explaining the poor schedule performance.
The Great Divorce: Cost Loaded Schedule Updating, John P. Orr, PSP,
ReplyDeleteThe article defines what resource loading means, cost, equipment, personnel, etc. It also outlines why you would want to resource load a CPM schedule.
Cost loading allows the owner and the contractor to anticipate expenditures across the length of the project. For the owner this provides a forecast of how much and when these payments are due. The owner may be able to cover the costs for a certain period and then need to start drawing against a loan. This would keep the financing costs to a minimum. For the contractor this provides a forecast of the how much they will be billing each pay period and how much will be billed to them.
The author spends a great deal of time in this article discussing when a cost loaded schedule is used to justify an application for payment and there are funds withheld for various reasons. If funds are withheld and the line items is not allowed to progress to 100%, the successor items cannot progress to 100% until the predecessor is 1005” This scenario forces the CPM to be artificially adjusted, illustrating incorrect work progress and funding.
The author suggests to “divorce” the cost loading and the CPM. This will allow the progress of the work item to be 100% while allowing the costs to be withheld. Separating the two makes some sense after the initial cost loaded CPM is submitted and approved. When the project is interrupted by an event causing a major cost issue, the project should be recalculated with another cost loaded CPM in order to illustrate the recovery o the project. When an issue is minor and does not really hold up the successor items, the funds could be withheld using a divorced schedule.
Having said all of that, isn’t this what retainage if for? Funds withheld as an incentive to complete the work on normal rework, corrective work issues. It has been my experience that major issues force a reworking of the project schedule and finding.
The Great Divorce: Cost Loaded Schedule Updating: The point of this article is to show that although a cost and resource loaded baseline schedule is a great planning tool during the course of a project, it may distort the CPM schedule and should therefore be kept separate.
ReplyDeleteUnlike a dedicated CPM schedule, a cost and resource loaded CPM network schedule includes a basis to measure: time, cost, earned value, resources, risk, information & communication, coordination, work organization, forecasting, financial planning, record keeping, and dispute resolution – no wonder it can distort the basic CPM timeline. The cost and resource loading requires the preparation of the Work Breakdown Structure which is in itself always a good thing; and all the other data that can be generated for other reports can be good as well, however, his main point is that many of these project management tools have competing purposes and are may be best when not put in a format where they are forced to try and integrate.
His argument (which I agree with) is that the primary purpose of the CPM schedule is to plan and manage the work flow of a project and aid management in making decisions based on how things are going during the construction phase. That is to say, the schedules main purpose is in planning ahead for material & equipment deliveries, coordinating the various subs and work assembly, and adapting in the field and making schedule changes to optimize the project time line, based on the realities of the actual project progress.
In his first example, Mr. Orr shows how a schedule with a single resource (workers) added to the mix, can distort the completion date if the software is allowed to “level” the resource. Essentially, if the software shows that the maximum number of workers on site on a given day is exceeded, it may level those workers out by stretching out the timeline. In reality, this may not be the best solution or be required per the situation. Cost loaded schedules may also distort the true schedule in that being 97% complete from a cost standpoint may not have an impact on the actual time schedule, but the cost component in the software can make the timeline murkier. From a cost perspective, the job is not complete and payment in full cannot be made, but this may not have any effect at all on the actual CPM schedule. The software however may determine that this does have an effect on the schedule if it determines these are predecessor components and the next activity cannot be started until the predecessor is complete. I supposed the old adage; “too many cooks spoil the soup” could apply, in that if software is trying to hit too many targets, it will be less accurate in its primary purpose.
In response to the “Great Divorce” Cost Loading Schedule Updating:
ReplyDeleteOverall I found this article to be very informative. The author stated in the article that a CPM Schedule serve more than just a place holder and calendar to mark the progress of activities. It allows for actual cost to be compared against budgeted cost. Also made apparent are those situations where time and dollars are to unlinked or “divorced” as the author puts it. By “divorcing” the cost percentage from the schedule percentage an accurate depiction the schedule can be obtained while keeping track of predecessors that are not complete but are not affecting the starting of their successors.
The response to incomplete/unsatisfactory work the owner may withhold just less than 100% of the payment in order to ensure the work is fished and not overlooked. In doing so the substantially completed work does not match up the cost loaded activity which will show out of sequence work that can cause the overall completion, earned value, cash flow projections etc. to be in accurate.
Out of sequence reporting can be carried out three different ways by Primavera. Retained logic, progress override, and actual date. Retained logic is the most commonly used and allows the start of the successors but can be completed until the predecessor of the activity is completed.
Out of sequence work can corrected by using the progress override but in some project contracts is strictly prohibited due to:
It automatically disconnects the original schedule logic for all out-of-sequence work disregarding whether the original logic was accurate and should continue to impact the CPM calculations.
It disconnects the uncompleted predecessors from the schedule allowing them to finish any time prior to project completion if they no longer affect their started successors.
Resource loading can also track the progress and help the project management team forecast earned value as well as completion dates on activities. By knowing the amount of man power that can be allocated to the project by certain contractors every man hour can be effectively allocated and evenly distributed without exceeding the resources available.
Just wondering what is The raison d’etre?
In response to Labor resource planning for construction project:
ReplyDeleteWell for starters I couldn’t agree more about the low unemployment rate and lack of people entering into the construction industry. I believe a better analysis would be that people who are in the industry are getting out if possible and the few that remain are hanging on by the skin of their teeth. I feel as though due to the economy we are faced with in today’s time the few good paying jobs available are not enough to warrant future generations making an investment to gain knowledge, tools, and learn leading edge technology. Sooner or later the construction industry has to bounce back and with the lack of new incoming skilled laborers the construction industry will struggle. Let’s face it the senior level experienced labor isn’t getting any younger and will not be around for ever. The new entrants into the industry will have to step and take their place. Skilled workers are a must on a construction project and unskilled workers when used to cover a shortage of skilled workers can cause more damage than good. Training of the new incoming entrants into the construction industry should be pushed to the limit and should be required that optimal training be given by senior (skilled) workers. In doing so the shortage of skilled workers can possibly be minimized.
I was thinking retainage (retention?) as well but decided not to raise the issue in case I was missing something. I'll be curious to find out if there is a difference between divorcing cost loading and CPM and retainage?
ReplyDeleteDS
Labor Resource Planning for Construction Projects, Bhargava, Oztemir and Dr. Chasey 2002
ReplyDeleteThe article proposes a model for the accurate estimating of resource needs (skilled labor) during the design / planning phase of a project. The research was conducted specifically on semiconductor manufacturing facilities. However, the authors suggest there is a need for further study on other types of projects. The continued research was undertaken by the Construction Industry Institute (CII).
The research is attempting to illustrate the effect of a shortage of skilled labor on the production and cost of a construction project. Inter views were conducted with industry professionals to gather the data for the study. A “base-build” was model was constructed using definitions for skilled labor. The collected data was compared to a recently completed facility of similar size in an attempt to validate the “base-build” model.
The comparison indicates the “base-build” model is fairly accurate.
Model Similar Facility
Peak Labor 1,710 2,050
Total MH 14,683 17,764
Area 393,800 483,000
MH/SQ FT 8.9 8.8
Total percentages of skilled labor
Model Similar Facility
Electrical 24 25
Mechanical 30 31
Masonry 15 17
This study should be completed for all project types but the information learned is very encouraging. Knowing an approximate number of man hours early in the project would be very useful in planning the construction schedule during the planning and design phase.
Artificial Resource Loading for Schedule Review, Albert Joseph Kuhn, PSP 2007
ReplyDeleteThis was a very confusing article for someone without experience in producing a project schedule in one of the softwares (Primavera 3.1 or Microsoft Project 3000) cited. The author gives very detailed steps to us in each softwares. You will need a working knowledge of these softwares in order to fully understand the proposed process. I gave up about half way through the Primavera discussion.
Labor Resource Planning for Construction Projects: The authors believe they have created a model which can help prevent or reduce project delays attributed to labor. Although this article was written in 2002, it is going to be extremely “current” as soon as the US housing market shows any increase in signs of life.
ReplyDeleteAccording to the authors, current economic conditions notwithstanding, part of the decline in availability of good skilled labor, is one of image. Another is the cyclical nature of construction. Construction is generally local and if there is little work, the labor leaves or enters other areas of employment. From my perspective, the overwhelming dominant factor is the cyclical nature of the work. There is very little stability in the work, and tradesman cannot and do not sit and wait for construction work to pick up if they have been laid off, but turn to other means to support themselves and their families. When construction picks up, there may not be enough labor to fill the void. This in turn can and does create delays in construction projects.
The authors suggestion that one way to approach this dilemma “could be the better utilization of training resources” and “improved planning, management, and supervision could result in actions which better match the number of new industry entrants to industry needs”. I don’t feel this actually addresses the problem at all, which is the inability to forecast the future construction demand. If the forecasting could be done accurately on a micro scale, then the above mentioned solutions would help the bottom line but they don’t solve the problem of unreliable forecasts.
Their example using the construction of a design build project of a semiconductor manufacturing facility, illustrates how lack of skilled labor caused delays and actually increased labor costs because of a lack of skilled labor (especially in the mechanical and electrical trades). In effect the project labor problems were inflated by overloading of unskilled labor to try and compensate for lack of skilled labor.
I believe this model is a good tool in showing the effects of trying to compensate for the lack of skilled labor with additional unskilled labor is unproductive and costly but really has no benefit or relationship to solving the actual problem of skilled labor shortages in a cyclical labor industry.