(1) Explain your problem, don't simply post "This isn't working". What were you doing when you faced the problem? What have you tried to resolve - did you look for a solution using "Search" ? Has it happened just once or several times?
(2) It's also good to get feedback when a solution is found, return to the original post to explain how it was resolved so that more people can also use the results.
Hi, Please, correct me where I'm wrong.
I remember reading on some SDPM paper that the schedule updates should be made daily to calculate the success probability trends ¿Did I actually read that or am I just imagining it?
Now, this "Daily update" is a hard thing. How can you update daily a schedule for, say, an hydroelectric project? Given the enormous amuount of resources working at field on the same day, I consider this a very hard task. I mean, from my little experience, it is just an ideal: they gather field reports to know "how is the project going", but usually these reports don't go to the schedule (I am speaking of Mexico), the schedule is actually rarely updated. When updated, it is made based on percentages dicted by field managers "We are 15% ahead, but don't ask me how much have I expend. I need time to gather the information and give you an estimate"
What I'm trying to know is, what is the required effort to actually track objective data from field to feed the project database? Can this be done daily? Is there some project doing it?
Best regards,
Carlos.
Printer-friendly version
Send to friend
Vladimir,
I know but it is time consuming and wonder if time cards will make it easier to enter details as per following figure.
Work on the activity varied each day and although having a Monday to Friday calendar some work was performed on a Saturday. If you use regular calendars to determine true actual duration using s single value of start and finish the results will not reflect reality.
I can consider some partial work as partial assignments and for analysis the activity was worked the full day but must account also for the week ends when taking out delays by Owner using a collapsible as-built, the activity duration must account it worked on a Saturday. Say it worked five regular days plus a Saturday, the duration was 6 days, if I model the same activity a different time using a five days per week calendar it will account for the fact it takes six days, perhaps a week from Monday to Friday plus next Monday. Of course the protocol use a trick latter on and suggest using 7 days week, a single calendar. It is about forensic needs being different than the modeling needs to manage your job.
Being able to update with such detail can be of much help for my needs, even if after the facts. Just show the facts then make the adjustments for your forensic model be valid. Assume I have no problem with identifying what duration is the net of a partial assigment, what duration is an owner delay and which is a contractor delay pacing is also to be considered can be owner as well as contractors pacing.
Best regards,
Rafael
Rafael,
in Spider people enter what was done during specified period:
what activities were executed and for each activity what volumes were done, what actual work hours were spent and when the work at this period started and finished? what resources were involved and what time each of them worked.
If the period was large and work started, stoped, then started again it is necessary to enter and submit actuals several times (for each period of continius work). If it is controlled we recommend to make the periods for entering actuals shorter (each day instead of each week for an example).
If it is sufficient to know only start and finsh of each activity execution during monitoring period it does not matter. You may enter that activity execution started on Tuesday and finished on Friday but during this period it was done only for 6 hours.
For me it represents no problem at all I wonder about how time cards will be posted and how each work hour will be identified as work being performed [or not] on the activity.
Reminds me of an issue I had decades ago trying to develop time cards with a database [believe it was dbase for Mac] that would allow for multiple rows of entries on the same record. Every employee weekly time card would be entered into a single two-dimensional array, a line for each cost code and a column for the hours worked on each day on this code.
By the way at that time Great Plains Accounting - Job Costing Module was excellent but it disappeared and eventually became mere budgeting when IBM acquired it, with no actual production considered [no volume of work considered], just costs and a % complete. Today when I look at systems like Timberline Job Costing I wonder if it is job costing or mere activity budgeting. As Carlos said paper can hold anything you write on it.
Yes, if you want to add several periods of activity performance you shall enter and submit them one after another.
it is possible to enter actual data on the same period several times
Of course but not on the same actual data input table, you must transfer one record at a time for a single activity. Take for example an activity worked 4 hours on Monday and no work performed until Friday.
For good record keeping and some forensic analysis you shall keep record of all interruptions otherwise goes against the contractor claim while leaving out work on some weekends goes against the Owner. With a method I am using the AACEI RP requires you to record all work done during weekends as this is activity time you shall keep when you take out the delays by owner. On the other hand if there are work-days in between you might be able to claim it was pacing and take out these chunks in your analysis, of course the other side will contest your claim to be a contractor delay so it is better not to abuse it.
Perhaps I am missing something about activity updating?
Best regards,
Rafael
Rafael,
it is possible to enter actual data on the same period several times.
Android application development has no highest priority at the moment but it may become necessary soon.
Vladimir,
You got it right about entering the data.
About Android application I do not believe you will find no more than a handful with such a need in my neighborhood. But never say never, better have such an option in case you need it.
I wonder what would happen if by error you enter twice the same time card, at least the software should warn and let the user decide. Guess time card can be for labor as well as for equipment.
Our traditional Job Costing applications include not only a Time Card but also the missing leg, the production report, otherwise it is only budgeting. Time cards represent an alternate way to enter actual resource hours along with activity dates but might be good to have an alternate way to enter production that follows the ways of the industry.
The scheduler will always be needed to update activities by others not in your time cards nor your own production reports, just much of the dirty work can be mechanized when working on enterprises that use those systems. If not difficult to program then why not, the performance archive shall be able to handle anything you throw at it.
Best regards,
Rafael
Carlos,
yes, we are thinking about creating monitoring tables for android. But today they can be easily exported to Excel, filled in Excel and imported back. Entering actual data in Excel is more complicated but possible using any OS.
Rafael,
I did not understand your proposal about multiple entries of information about the same activity. Did you mean entering several times actual data about the same performance period?
Carlos,
If you are into productivity claims the following article can be of your interest.
Our lawyer have claimed 3 times lost productivity using Leonard and being awarded the 3 times. It does not means is a sure bet but a possible one. It is tricky and if you do the wrong math will be destroyed as fast as you can assemble it.
Is similar in a way to Echelay and perhaps easier. After the determination on lost productivity factor to use, a single percent is applied to the total man-hours of original work by contractor to get total man hour lost. Then an average cost per man-hour is applied. I am involved into other parts of the claim and trying to leave the details of this computation to other engineer that after a brief discussion shall be able to handle the math. I made a trial run using guess-timates for the total man-hour instead of the job payroll records, and the percentage using Leonard and Gibbs for the project conditions we found out that difference was a huge 0.0000000000000000000000%, no matter what the final number of total man-hours the timing of the changes (late) considered by Gibbs make both formulas to yield same percentage, but under other conditions they can differ. We believe here we have a good shot as they converge for the specific job conditions.
About Spider Monitoring Table maybe it shall be adapted to accept multiple entries on the same activity as the workaround to this issue I use is entering and posting multiple entries to the same activity by entering and posting one by one. You cannot enter multilpe entries to the same activity before posting. It is tricky but I am sure Spider team if there is demand for it can implement some time card entry, maybe there is one already. The good thing is the math is always correct and Spider reports do not depend on financial periods. Maybe there is a better direct way using Reference Books functionality as this sutuff let you manage the database without risk of damaging the model. Because I do not have a need for it I have never asked for such functionality.
I am having fun with Spider in the preparation of our claim as just the links table makes a whole world of difference, and this is just a table not about true Spider functionality. In our claim the original schedule was developed using SureTrak by other scheduler and all the functionality of SureTrak is a small subset of what Spider have. More fun will be when using advanced Spider functionalities no other software will be able to mimic in any way. The issue is getting a contrator delayed by Owner when using Spider who want my services.
Will keep in touch.
Regards,
Rafael
Thank you all for your opinions, I consider them very valuable.
Anning, I know that engineering is an ideal but work on site is very different from "paper" engineering. We have another saying here in Mexico: "El papel claro que soporta todo, pero intentalo en realidad y me vienes a contar" "Paper sure will support everything, but test it in reality and tell me" I have similar experience to yours, I was working at an hydroelectric project and know "What it feels". Even with bad management, there are many hydroelectric power plants built and running. The part where I don't like it is when these huge projects ends with claims. I think that daily update is not an "impossible" but more a "challenge". I am concerned, in your opinion, why do you cite these cases as non-daily updateable? Specially the diversion tunnels, from my experience that was the only part of the project which got closer to a daily tracking. The SDPM paper may sound unrealistic, but don't you think it's a great ideal to pursue?
Rafael, I will consider your method as a first goal. The company where I am working now is not very interested in schedule tracking but more on productivity tracking (because our business is construction claims and lost productivity is usually an unexpected and succesful claim) {Disclaimer: I don't consider construction claims always ethical but I am here for the times when it is ethic}. I think that productivity tracking comes in hand with schedule tracking because the schedule can be loaded with resources, thus teling us what to expect. I find your approach very practical, though, and a good practice that I will have to apply when I start to track a construction schedule again. I think it will work towards pursuing Vladimir's daily-weekly update ideal.
Vladimir, thanks for the advice. The company where I am now is a little more open to change, but not involved directly in construction but in claims. Still, everyone here seems concerned that the productivity reports that we get from field are not reliable, and we are making an effort to know how to make those productivity reports more reliable. Tell me, have you thought of developing the Spider Monitoring Table to be tracked on Android or iOS devices? I think that the main difficulty to track productivity information is from field to management, what do you think?
Best Regards,
Carlos
Annung,
SDPM does not require daily update. This approach does not depend on the frequency of project updates.
But still we kept weekly updates in the project similar to yours http://en.wikipedia.org/wiki/Boguchany_Dam
Carlos, my opinioni is, and i agree with you completely with you that daily update for a mega project like hydroelectric project is a hard thing. Why? From experience.
In a 2400MW Hydroelectric projectin FE Asia, we had about 35,000 activities. Our Design Team is about 2000 KM away in another state (for design updates), the procurement team is also that far. In the remote region, where there is always heavy rain (hydroelectric projects are generally located where there is high amount of rainfall), the internet connection is not always available, and design and procurement updates is not always available.
What about 30+ subcontractors, (large and small) who have their daily problems; even for their own subcontractor, it is difficult for them to assess their progress, what update info can be available for the main contractor fo daily update?
Let us site a few example of activities that can not be update daily for a hydroelectric project?
1) Excavation works to the Main Dam, Powerhouse, the Spillway, Power Intake channels - (here you need the daily trip records of each truck from day shift and night shift).. unless the update will just guess estimate.
2) MEP Daily Updates for the installation of power house and the power intake channel.
3) Diversion Channel tunneling activities
The SDPM paper is good, but may not be realistic.
Like in construction, we ask, maybe the design looks good but is it constructable?
Carlos,
I agree 100% with Vladimir, but I have my own variation when doing schedules on my own jobs.
We manage an official monthly update that will include not only actual dates but other data such as resource usage and costs. Definitively waiting for the end of the month is too much time. I create a copy of my monthly updates and from these I create my own weekly updates for purpose of managing the daily operations. These updates do not require historical resource or cost data. This is for exclusive purpose of looking ahead. Remember that when looking forward not every actual date before DD is irrelevant as some some links can still be active, so try to be as close as real with your dates and estimated remaining durations.
For the requirement to keep true records we use the monthly update. These monthly updates are very important in the case of a claim you need to use the data in a forensic way. Latter you might find merely using a single start and single finish date for your activities is not enough, forensic analysis require you consider work on week-ends as well as non work days in the determination of As-Built durations. If you are not entering such detail then make sure your daily reports are good enough to identify and cross reference to CPM activities.
If during the first week of the month there is a change in some critical delivery are you gong to wait 3 weeks to see the full impact and react if need be?
Hope this idea can be a compromise you and your company can adopt.
Best regards,
Rafael
Hi Carlos,
usually construction schedules are updated weekly.
Updating is usually linked with project management team meeting. Project planner(s) prepare project status and trend analysis to this meeting where corrective actions are discussed and approved.
It is not rare when updating has different timing at different levels of project management. Weekly updates at the top, daily updates on sites.
Spider Project Professional replicates project schedule sending subproject models to people responsible for their execution. At the top project model is updated weekly collecting latest information of all subprojects. But subprojects can be updated daily and managed locally without data exchange with central model except once in a week.
I would recommend to start with weekly updates on all management levels. Field reports may be collected in table form (Spider monitoring table) that contains the list of all activities planned to be executed during considered period, planned volumes of work to be done and planned durations of planned activities. Once in a week people shall enter what was really done in physical units and not in percents. Spider will collect this information and update the schedule automatically. It is not hard when people have nothing to hide.
This routine that I described is used in many projects including Boguchansk hydroelectric project. There construction schedule is updated weekly.
Email: PPadmin@planningplanet.com
Replies