Esko,

We have a document that lists Spider Project fields it is a 44 pages document with an average of 50 fields per pages, all these files are necessary to transfer all data within Spider Project, about 2200.

Each software has a unique set of relational tables and many values have no equivalent on the other software. We know transfer of data tables will never be 100% and in addition, each software manipulates the data in different ways.

Someone has to decide on the standard recognizing depending on your software some of it will be read, some will be discarded and that even when 100% can be transferred, something I doubt, then each software might yield different results.

P6 users say they have many % complete definitions, but we do not use them, we use the actual planned or realized values as % complete can create issues with loss of significant digits. We can create our own fields to define hundreds of % complete fields and with formula manipulate some data, I would never go this way though, we prefer to keep the precision.

P6 has financial periods, we monitor performance in a different way, we simply have no need for predefined financial period and can report on any period to the hour, can be starting from any day of the week at a particular hour and ending on whatever date we choose, we do not average the data to latter on end up losing the details, we can define in our reports literally hundreds of financial periods without manipulating the source data, the software will do the math with precision. Here we have a monumental difference between data tables.

Imagine SDEF as a wharehouse of parts for cars, If I have a Toyota and the SDEF has a Subaru transmission it will not fit into my car, same might happen among the software, a few fields can truly be shared, like gasoline, water, motor oil, perhaps the tires but in no way the cars will run the same, the ride on each car will be different.

SDEF shall be targeted as to keep a record of some basic data but we cannot pretend it will result in perfect transfer of data, in their ignorance some people perhaps were expecting too much, a perfect transfer of data and some magic to make all software perform the same. SDEF shall transfer the dates to keep them on a database for the record, to be used only if need be or to transfer fixed dates into a Portfolio, data fixed in time and not to be manipulated by other software.

If you want the data to be manipulated because you are the Contractor or a PM if work is done under a PM Contract then you will need the data to be transferred using a particular conversion utility for each version of source and target format. After this initial transfer the data will still have to be manipulated by the other software in order to get a close model, a not so close model I would call it “not even close”. The best thing you can do is get to “not even close” after a substantial amount of manual work. This is intended to be a one time operation by experienced planner and scheduler.

I believe the only way to solve the issue on being able to see what my software shows is through the use of free viewers that will allow hundreds or even thousands of people take a look at your data, without being able to manipulate it. You shall be able to print your own reports using the viewer. If you want to manipulate the data, the same as the original software, then buy yourself a license of the software.

Still for your records and for some Portfolio Management SDEF is the way to go. SDEF shall provide for you to do data validation and to perform some forensic analysis in case there is need for it, something shall be very rare but in ocassions very relevant. Our agencies must manage thir portfolios and must be able to have the appropiate records in electronic form.

Fortunately, we have Agencies such as the Corp of Engineers where the standard is high. My l experience with the Corps of Engineers is limited to a single job, a water intake at a river. They provided the design of a fish ladder to protect some of our local river fishes; a Marine Biologist from the Corps would follow up on the construction of the ladder and believe they even provided a percentage of the funds for the ladder. The Corps have been involved here in some water reservoirs and everyone respect their professionalism, is top notch.

Regards,

Rafael

Esko,

I agree with you and believe the SDEF Standard is the way to go, even when already obsolete it still can do the job for Portfolio Data. I believe it was the Government who discontinued the use of SDEF on purpose, in collusion with some vendors. Yes an updated SDEF is overdue, even when it will never make different software yield exactly the same results, or even communicate effectively. For example we can define a unique calendar for every lag, independent of a limited choice across the board, we have temas that make it possible to model correctly shift work but P6 do not, therefore only on team is transferred to P6, we have multi-resources and partial assignments, P6 does not, and the list goes on. That is why a free viewer is a must have in order for others to have a complete and unobstructed view of your schedule.

Because Government do not require SDEF most developers abandoned the support for SDEF, and Primavera got to be happy with it, perhaps they still support SDEF as a strategy to avoid being blamed on what in reality is not their fault but the fault of governments, Federal, State and Municipal.

About the Corps of Engineers I must take my hat off and salute them, is a pity other agencies did not followed their example, starting by the GSA. Yes, we still got many good people in our government; the Coorps of Engineers is an example of how good our agencies can be.

Regards,

Rafael

Esko,

Whatever the excuse it does not justifies a monopoly. There are still many people in government that believe the ends justify the means. It happens in our military, in our Federal, State and Municipal level, but that a procurement agency such as GSA openly disregards good government practices is abusive.

Probably the government abandoned the idea of such specification because it is easier for them to specify single software at the expense of the Contractors and other software vendors. They do not care at all, if you got a government job you have to kiss their butt and thank them in hope they do not retaliate and get the jobs to other jurisdictions.

The problem with exporting to P6 or any other standard specification is that during the process a substantial amount of fields are not transferred, essential data fields needed for more advanced resource leveling and loading than what neither the specification or P6 can handle. You end up with a different model. By simply requiring electronic files and a free viewer capable of displaying all fields and logic the problem can be tamed.

For Portfolio Management the projected dates shall be enough, this can be transferred by use of the specification you are talking about, even the Flintstones Version will do it.

Asta Power Project and Spider Project provide for such free viewer, perhaps P6 shall be required to provide a free viewer.

Regards,

Rafael

Because usual confidentiality requirements I will not post the Specifications but will post my comments to GAO as follows.

COMMENT NO 01

The use of an independent CPM consultant is contrary to other agencies policies as to promote the use of in-house personnel for the creation of the CPM Schedules. Today Project Management courses with emphasis on its application in the Construction Industry are frequently taught at most Civil Engineering Colleges. Usually contractors have qualified personnel with engineering degrees, with the experience into the planning of the types of jobs that the contractor does. The requirement to use external scheduler promotes low Project Management culture among the contractors.

People experienced in the type of work to be performed shall do the planning; planning shall not be done by an external scheduler, at times a keyboard jockey that has typed a few similar jobs for a few contractors, this do not qualify him as a planner, perhaps as a data entry clerk. The experience requirement shall rest on the qualifications of the General Contractor assigned Project Manager, the person responsible and with the authority to do the planning.

Scheduling is a tool for planning, not otherwise.

Project schedule shall be developed taking into account Contractor resources capabilities and availabilities, internal production norms and resource crews. It may take more time to teach external consultant on Company specifics than to create the schedule by trained internal team. In any case if to involve external consultants or not shall be Contractor business decision. The Client may insist on certain requirements to the schedule model but who created the satisfactory model shall not matter much. The requirement that Consultant shall not have any financial or business ties is absurd. Construction Schedule is usually created by project management team that includes project manager, project planner, project estimator, procurement specialist, cost engineer, and site managers, is a team effort. Including external Consultant in this team may be useful but only Contractor shall decide if external help may be useful.

Construction schedules shall be created basing on the quantities of work to be done with the application of the corporate databases like production norms, norms of material requirements per job physical unit (volume unit), etc. Primavera does not work with physical quantities of work and norms. Therefore, we will not be able to collect actual data on the work done (in physical units) and enter it in the scheduling tool for automatic rescheduling of the remaining works. Primavera is based on time estimates that may be sufficient for non-demanding Client but not for the Contractor.

COMMENT NO 02

The qualifications requirements are not specific, I suppose under such ambiguous specification the interpretation is whatever is reasonable to be interpreted by the one who did not wrote the contract, a usual interpretation of the law.

In the same way brand specification shall not be done, the qualification requirements shall not be discriminatory to our professionals trained in Puerto Rico in Colleagues and Universities accredited by recognized institutions like the Council for Higher Education Accreditation. Perhaps accredited technical institutes can be acceptable but not to exclude our institutions of higher learning.

Once again, the experience requirement shall rest on the qualifications of the General Contractor assigned Project Manager, the person responsible and with the authority to do the planning.

“We are the only who can qualified you” It is kind of a snobbish attitude that can leads to unjust rejection or even worse, favoritism to a particular brand and service suppliers.

I did not see any specific qualification requirements. But presenting the schedules of other projects is unbelievable. Here construction companies will blame the Consultant for disclosure of their confidential information. Construction schedules are usually based on internal norms and cost estimates that are confidential. It is usual to sign non-disclosure agreement with persons that are involved in construction planning.

In my professional practice, I consider the non-disclosure agreement signed by my clients as if signed by myself, after a job is finished all paper and electronic files are delivered to the client and I keep no copy as a statement of my commitment not to disclose confidential information.

In any case limiting the schedule preparation to an External Consultant is a bad practice and foolish requirement. Forcing the contractor to change his planning tool, on every job, as if an underwear change is also bad practice and foolish requirement.

COMMENT NO 03

The repetitious use of the Primavera brand speaks by itself, this is a brand name specification intended to rule out all others. Is against the most basic procurement policies intended to promote competition and to get good value for public money expenditures.

COMMENT NO 04

The specified format for the Complete Network Diagram is impractical for the required schedule size, required perhaps because of the inability of Primavera Products to generate better ways to follow the network logic. Yes, they create impressive graphics, perhaps good to be displayed as if a Picasso, but useless to follow when the schedule is of several thousand activities. I cannot imagine anyone who would use such diagrams for complex jobs as thousands of links spanning the multiple pages will become impossible to track. Remember thousand activities means at least twice the thousand number of activity links.The importance of being able to follow network logic is not in question, but shall be performed using practical methods.

Although useless for most practical purposes these diagrams are available on most commercial software, the requirement should be for a better report and better computer layouts that will help you to audit the logic with the aid of a computer, PERT trace logic available in most software still falls short, is inefficient. These diagrams are limited as to how much of the information can be displayed, particularly links related fields. However, Primavera software lacks good layout views for this purpose as available in more advanced software. Some Primavera reports can make up for the lack of these views.

If on average you can squeeze the diagram boxes as for each have a tributary area of 3inx3in then 3500 activities would require 31,500 square inches of area, the equivalent of about 25 sheets of the required size 30x42 inches. Because the geometry of the network is irregular this estimate fall short but a gross estimate could be that a surface as big as a tennis court would be required to display all sheets, scotch taped on common boundaries, of course the use of additional empty sheets can be waived, just print those containing the little boxes.

For a 3,500 activities network estimating that an average of 3% or 105 activities fall into the critical path perhaps you can estimate twice this amount as to the quantity of columns required to display the network grid. This is because in between nodes a couple of critical activities might occur in parallel with a larger number of activities, also in tandem but of shorter duration as not to make the path critical. This would make the grid to be of 210 columns, at 3in/column will require a 52ft long display.

I wonder if an IMAX theatre wall would be enough to display such a diagram for a larger job, say a 60,000 activities job.

COMMENT NO 05

The requirement for number of activities is a wrong method to specify required level of detail; it is better specified as to maximum activity length and other requisites as already required in the specs. The activity count shall be as many as required, if 3500 is not enough then it is not enough but if 2543 is enough then it is enough.


COMMENT NO 06

The requirements for how risk analysis is to be performed is not disclosed, is ambiguous leaves it opens to the caprices of the VA, maybe only as per the Primavera ways, even when other software can provide such analysis in way that is more efficient.

The specs requires performing risk analysis using additional software as Primavera P6 is by itself off the shelf is incapable of performing such analysis. They usually export the schedule to separate software known as Primavera Pertmasterto perform what is known as a Monte Carlo Simulation. However,Premaster uses a different scheduling algorithm than P6 so the modeling from the start is in bias. Of course, no single project has statistical analysis of hundreds of similar jobs for the probabilistic distribution being other than a “foolish” guess. It does not mean statistical analysis is useless, just that it is not as precise as people believe or pretend. Are they to suggest you use a standard distribution across the board for all activities? Under what reasoning?

The Contractor must make sure his scheduler qualifies to perform such analysis, whatever qualified means, they should be clear prior to bid opening what the qualifications are, otherwise who is to interpret the requirement?I suppose under such specs the interpretation is whatever is reasonable to be interpreted by the one who did not wrote the contract, a usual interpretation of the law.

Perhaps they want to disqualify non-Primavera products with this requirement but there is more sophisticated software than Primavera commercially available at lower cost. They are promoting a low Project Management culture or perhaps none culture at all, they want contractors and everyone to believe contractors are incapable of performing their own scheduling with their own people. If the government has a need to manage their portfolio of jobs they can do that merely by requiring the information to be submitted in a standard file format.

Excluding other software vendors is like asking a contractor to use exclusively Caterpillar brand when most of us have Komatsu. I have a Toyota 4Runner, can I drive to the meeting on it or it must be a Ford Pinto?

Besides, Risk Analysis is not only mitigation plans. Using risk simulation we will get the probabilistic distributions for project duration and cost. To make project targets reliable we shall create and manage contingency reserves. It means that working schedule shall be tight but the reserve shall belong to Contractor. In other case, risk simulation is useless. Float consumed by the Owner reduces the Contractor contingency and must be replaced with EOT. If the owner pretends to share ownership of float then once again, risk simulation is useless.


COMMENT NO 07

Among the remedial actions called for is the increase of construction manpower and the use of shifts. However, primavera is incapable of modeling the creation of resources and true modeling of shifts.

Although I do not have P6 as I do not use it was disclosed on a Planning Forum that.

1) P6 does not provide for the creation of resources; only use fixed dates to determine the availability of resources. This is not how the Contractor plans his work force; we determine reasonable levels of work force depending on the stages of work. During peak concreting phase we keep a certain level of manpower while thereafter the availability is reduced, otherwise idle resource costs would drain our profits, if any. On many thousands activities job with hundreds of resources types manually adjusting the availability is impractical.

2) P6 cannot effectively model shift work.

“P6 models shift work only if you know before scheduling what work will be done in what shift. Of course, it is not known. For example, imagine some work that has 6 hours duration. It may be done in first shift, second shift, or one part in the first shift and another in the second shift. In Spider Project, you can assign both shifts and the work will be done by those resources that are available. In P6 you shall specify what work will be done in what shift. So an activity will be delayed until assigned resources will become available.”

I would add that Primavera P6 resource leveling algorithm is inferior to many other available out there. Even NASA uses an add-on called Aurora to use it as the engine. On their site with a sample schedule Aurora demonstrated the superiority of Aurora over P6 with regard to resource leveling. I modeled the same sample job using my software and got better results than Aurora who got better results than P6. Even with MS Project, allowing for activity splitting I got better results.

http://www.stottlerhenke.com/products/aurora/Turnaround/2009-10-01_Auror...

From the above link:

“The goal of resource leveling in Primavera P6 is to provide the user with a valid resource loaded schedule that does not have any over allocated resources. Primavera P6 does not try to optimize the allocation of resources in order to generate the shortest resource leveled schedule. Even though Primavera does not purport to providing and optimize schedule, is likely that many users of Primavera P6 are NOT aware that the results from the resource leveling process are not optimal, and could be improved upon significantly.”

How can it be possible we are denied the opportunity to use a better tool? Is it a non-written government policy to require all the exclusive use Primavera products?

COMMENT NO 08

The limitation on the amount of characters to be used on the activities ID combined with the limitation on Activity Description will make it almost impossible to navigate on the activity drop down menus to select particular activities. This will make it harder than necessary to create a numbering system for organizing and grouping of activities, it is common practice to use letter prefix in addition to numbers in order to do so. It is not just about the organization of views and reports but also about the creation of filters.

A FEW ADDITIONAL COMMENTS ABOUT THE SPECIFICATIONS AS ISSUED:

The specifications do not address the handling of the followings issues:

1- The specifications do not suggest creating contingency reserves. It means that any created and approved schedule will have low probability of successful completion. If the statistic results from the cumulative distribution curve for the predicted job duration are not considered in any way for the determination of an appropriate level of buffer then the statistical analysis is essentially a futile exercise. The management schedule must target for an early completion date in the hope it will finish on time. Some schedulers use Critical Chain theory but this lacks the methodology to determine appropriate level of buffer and induces artificial delays when apparent float is consumed and buffers enter into play, also resource-leveling algorithms are fooled.
2- The in-appropriate handling of rain days with the use of calendars for this purpose fools the early dates in a similar way Critical Chain and use of buffer does. The appropriate model would target for an early completion in the hope the job will finish on time, the possible effect of rain would be part of the risk analysis, and unfortunately, the specification is unclear and can lead to the use of wrong model.
3- The specification makes no reference for a standard procedure for Time Impact Analysis. There are several published standards, but which one?
4- The specification makes no reference to the use of a WBS or Work Breakdown Structure. If they insist in the use of P6 and a WBS that fits their particular need, with absolute disregard to our needs,the software will deny us to use our own. Primavera does provide for a single WBS dictionary, while my software of choice allows for multiple WBS definitions, this would allow satisfying the need of all.

In my 30 years experience as an engineer and Project Manager in multiple construction jobs I have never been confronted with such restrictive and biased specification for the use of a planning tool that shall be the contractor’s choice. Yes, I understand the need for some electronic transfer of data that will allow the Owner to manage their portfolios of jobs using the software of their choice, but it shall allow the Contractor to do the same, to use the software of his choice. There is no need for brand specification when the transfer can be using an open standardized file format.

A good technical specification shall be based upon the expert experience of many years, not improvised.

The same arguments I have against the use of P6 brand specification in government procurement stands for any software, I firmly believe that even my software choice shall not be named.

Best Regards,
Rafael Davila

For assignment and status information, please refer to the on-line Bid Protest Docket, accessible via the Legal Products link on GAO’s home page (www.gao.gov).

Vladimir,

I understand Primavera cannot not model shift work properly and now I can see at another thread it does not calculate resource floats correctly. I believe that the argument Primavera cannot do these two functions properly can add to why Primavera shall not be brand specified. That Primavera computations are wrong can be a good argument against its use.

Can you provide me with a concise explanation for each case of why it cannot?

Best Regards,
Rafael

I tried the following with no response to let me know if the communication was received and if I can expect some type of action or not, not response whatsoever to tell me my case has or no any chance. Any other ideas will be welcomed.





For your knowledge GSA issues with software suppliers is not a new thing. As I see it Monopolies are good business, the problem is when it reverts against your choices.



Well being sarchastic is all what is left but I am not giving up, even if it takes a lifetime.

From:

http://www.cecer.army.mil/techreports/Eas_fix/Eas_fixb.pdf



It is obvious there is no need for proprietary scheduling software specification other than stubbornness of the government officials who insist in insubordination or perhaps a conflict of interest due to possible corruption.

Best Regards,
Rafael