Meanwhile you can try resource leveling on activities whose duration depend on assigned resources productivities, with partial assigments so a single resource can work at the same day on more than one activity while modeling shift work for shifts that work on different days but same time of day.

http://www.mediafire.com/?64u0nubb1d8df

It is very very simple just two shifts on each activity, 3 activities. Just let me know if you was able to even model it in other software, not to say level it. To me the benchmark tests at which Spider is better than most or all of other software are not good enough to test real life resource loading needs.

You can download the free Spider Viewer and take a look at the 200 activities job but will not be able to edit the file, just view it and make reports.

Evgeny,

I suggest trying something a bit challenging like the following schedule you can download from the link

http://www.mediafire.com/?lcmrap5saaaicyj

It is still small only 200 activities, a single resource per activity, perhaps still too simple but a bit more than the simple schedule you provided. Spider have several algorithms for resource leveling, some depend on user input for prioritization rules and some do not, feel free to try different rules.

The following are the settings for Standard algorithm with Total Float as scheduling priority. You might notice some other software provide for several level of priorities in an attempt to improve their algorithms but improve just a little. Only by testing you can see the difference between Standard and Optimization and Optimization Plus.  

My approach usually is to plan my schedule with resources enough to finish ahead of time, with some buffer. Using standard algorithm at the baseline, then manage using "previous version" option that will keep activity sequencing. This algorithm not available in most other software will prevent chaos. When getting into trouble I look for other options to get my schedule back on track, at times working on Saturdays, at times adding a second shift, adding resources, using other leveling algorithm. Then on further updates will go back to "previous version".

If Asta does not provides optimization algorithms maybe you can get some improvement on the leveling by guessing prioritization rules, this applies to most resource leveling algorithms that depend on the user input of prioritization rules.

By playing with prioritization rules some people can get substantial improvement on simple jobs but on real life schedules with complex resource loading such as the use of skills and partial assignments it will be very difficult if not impossible to get substantial improvement over a computer solution.

Best regards,

Rafael

Such micro planning is best done at the monthly - weekly planning meeting.

There is a misconception about micro-planning as if it no longer matters, perhaps the stakeholders are not interested in the micro but those at the project level is what PMs do everyday. Even today mathematicians still believe in Newtonian calculus, the whole is the sum of many many micros.

The suggestion to resource level only for next week is in error, it will not warn you what might happen afterwards, frequently unpleasant surprises if you do not plan on time.

I would not advocate setting it down in the baseline.

You got to know what is the macro ahead not late, to get it right you need the micro. Anything after the baseline can be considered self-serving.

Even so I do suggest that each task has one trade (resource) on one location which can be leveled as suggested.

One task might have several crews, each crew composed of different trades but its members working together on each team. You can define each crew as a multi-resource of several skills, keeping the details and being able to load it at a single click of the mouse, we no longer use the Abacus. I guess that for claim purposes you must make it transparent what is the composition of a crew.

If you limit your activities to a single trade the software might resource level each trade activity that are required to work at the same time to happen at different times, wrong when the different trades must work together.

I advocate for the use of trades because of the needs of my schedules but others require going into the named resource details.

Partial allocations - such as plumbers - are then set onto a collection hammock - power point allows split resource allocations.

Plumbers are limited resources, probably one of the most limited. You have to resource level them into the appropriate activity. Loading a resource on a hammock of 100 activities will not tell you when 20 or 50 member activities are being worked together. Hammock is not good enough to model this resource allocation correctly.

Tower cranes are put onto summary bars.

The same goes for cranes as for plumbers, they can work on a limited number of activities each day and each activity might require different crane hours. Hammock and Summary Bars are not good enough to model this resource allocation correctly.

Because this is what happens everyday your software shall allows you to plan for it in an easy and transparent way for it to be practical. If you know your needs I do not see why it got to be difficult. It is easy, very easy, you know your needs. Allow the software to do the dirty work for you.

Here we allocate on a single formwork activity [out of many such activities] several crews:

1. a carpentry crew composed of a supervisor, several carpenters and several helpers all assigned usually 100% of their time to the activity.
2. a plumbing rough-in activity composed of a supervisor that checks several such crews at the same time but the plumbers work say 30% of their time on this activity, a resource on partial assignment within a crew on partial assignment.
3. the same goes for the electric rough-in crew
4. we also assign a crane and crane operator to work on several activities at the same time.
5. we also assign flying forms, gang wall forms, hand setting forms ... as limited resources, we do not forget about it an buy/rent limitless quantities of such resource. Of course as any other resource we might increase availability at some cost if need be.
6. It is not uncommon for resources to have different calendars, different vacation days.
7. It is not uncommon for activities to have different crews working at different shift hours.
8. It is not uncommon for activities [and also lag] to have different calendars.
9. It is not uncommon to assign resources on variable workloads so they can work on different hours on the same activity depending on availability.
10. Some resources must work together as a team while others can work independently.
11. and the list goes on. 

Nothing out of this world, just how real life resource allocation happens at the job level. The well known sample schedules used to compare and test resource leveling are less complicated than a mere fraction of 1% of real life scheduling.

Not always you need resource leveling that looks for shorter schedule but other algorithms that keep activities and phases priorities and most frequently algorithms that keep activity sequencing of prior versions. A single algorithm will not do it, you need your software to provide these and several other options.

I cannot imagine myself doing manual resource leveling in my jobs where we have several crews assigned to the same activity, each with different resource composition, each with different productivity and each with different resource partial assignments.

Don't even think it will be near optimal.

Even if your software is capable of displaying true resource leveling float values when you do not resource level then the float values will not be resource level values. It is ironic using CPM schedules for resource leveling without displaying true resource level floats. If your project is dependent on resource availability and you use manual leveling your float values will be unreliable, perhaps none will be valid.

Computers are here to help, do not work with 40 years old technology [substandard computer algorithms and very limited assignments functionality],  do not work with a few centuries old technology [using the Abacus]. Use your software resource leveling functionality, even if limited, on complicated real life models it will be better than human manual attempts.

Hi Evgeny

Welcome to the wonderful world of PowerProject

You may have picked up my earlier comments on using software to level resources - do not use it because it will screw up your programme in nanoseconds and you will never know what damage has been done.

Better still use other power project functions to achieve the same results but strictly under your control.

1. Use the Link Category to set up links for each resource - using different colours if you wish to show on the legend.

2. Filter to resource tasks and link FS using the relevant link - you can make global changes when all the links are selected.

3. Reschedule to level your resources.

These links can be switched on and off whenever you wish.

Also it is a good idea to set vertical links to a curve so that they are easily traced when printed.

Best regards

Mike Testro

Hi,

I'm afraid you won't find resource leveling as a function within Asta Easyplan.  That is our entry-level product which sells at a significantly lower price than both MS Project and Spider.

However, if you download our main product, Asta Powerproject, you will find a resource leveling wizard available under the 'allocation' tab of the ribbon toolbar.

Please let me know if you need any further information.

Ben @ Asta