How to erect a recovery schedule

A few random thoughts:

1. Nothing will happen quicker just because you reduce the duration of a schedule activity. Look at how long things have actually taken in the recent past and then use that time in the revised plan.

2. Identify the issues that have caused the delays (what is the root cause) and develop plans to remove the causes of delay. But be realistic - if the solution was simple it probably would have been implemented by now. 

3. Where you need to recover time, recognize this will need a change in performance - people will not simply work harder because you say-so....... Change in performance can be achieved by more efficient work practices, better equipment, more efficient supply-lines and/or deploying more resources but all of these need designing, planning and time spent implementing them BEFORE any improvement in performance will occur.

4. Remember Brookes Law - adding more resources to a bad project makes it worse! To be effective, before adding resources you need improvements in supervision, work practices, supplies, etc. 

And all of the above cost money. Effective recovery programs burn a lot of effort and money (slowing progress in the short term) before the benefits cut in.  For more on schedule compression see: https://mosaicprojects.com.au/WhitePapers/WP1059_Schedule_Compression.pdf

I'm completly agree with you, I think that it's better for us to make a recovery plan with the actual rate of production and always try to understand why we have a delay in the past, we must always make a weekly gaps analysis.

there are 2 basic things that you can do 

fast track or crash a schedule 

Fast Tracking the schedule means to Revise the logic and do thing concurrently or in parallel which means that your manpower peak will be higher but it will be the same number of man-hours except for additional supervision.A good rule of thumb is that sequential activities can sometimes be fast-tracked by up to 33%. In other words, if you're fast-tracking, you can start the second of two sequential activities when the first activity is 66% complete. So when you are 66% complete with the first conduit activity you can start the wire pulling.

Crashing the schedule means to reduce your durations but you will have to increase your man-hours because you need more men to do it in a shorter amount of time. You would shorten the duration of the activities on the critical path by throwing additional resources to the critical path without necessarily getting the highest level of efficiency. You might have to do several iterations because you crash one critical path then something else becomes critical. You will have increased cost in with this option

Always remember to check you successors especially the start to start with lags becasue if you shorten a duration of an activity but do not adjust the sucessor lag then there may no longer be an over lap and it may actually turn out to be a Finish to start because you did not adjust the lag.

Then there is using a combination of both of these techniques