[OT] - Estimating

Steve Etter

Lifetime Supporting Member + Moderator
Join Date
Apr 2002
Location
Morristown, TN
Posts
965
We have a machine that is currently down and being completely rebuilt. As part of that rebuild I have a sub-project to replace virtually all the controls, wiring, and conduit on the machine. It was known up-front that replacing my controls would be the critical path and that is no problem, however, the facility is in dire need of the machine and the folks upstairs are pushing for every possible expedite. Once again, no problem, I have pretty good plans, designs, conduit runs, wire-pull lists, etc. and I can get all the man-power I need to work 24/7 to get the job done. The problem I do have is my ability to estimate timing.

I am using Microsoft Project to work up my time-line and I have a very detailed list of most all the tasks that need to be accomplished. While I feel fairly confident that I have captured all the majority of items, I am not confident that I can adequately allocate, but not be too generous with the time required for each task. My worst areas are conduit runs and wire pulls/terminations.

The folks upstairs want me to forward them a copy of my time-line. Obviously, once I do that they will try to take out anything they think is fluff and ask boat-loads of what-if and how-come questions. These are no problem, but the one thing we all know they wonā€™t do is accept the finish date slipping simply because the time was estimated poorly.

So my question is this; how do you all determine how long any given conduit run or wire pull/termination will take? Does anyone have a spreadsheet or chart they use? Please remember this is an aggressive installation so I am looking for more realistic the data you might otherwise use for quoting purposes.

Thanks
Steve
 
you have to give more detail. you could run 400' of conduit in a straight line in very little time if its laid on the floor you might get 4' if its 100' in the air. on avg. 200' a day for 2 men is good.NECA has a labor unit guide you could use.there are lots of things to look at how many fittings,how many field i/o to hook up?
 
I am using Microsoft Project to work up my time-line...
I bet you are viewing your project in the "Gantt" chart view. Try the "Project" or network view. You may be surprised at how many loose ends you see hanging in mid-air, going nowhere, connecting to nothing, with either no start date or no end date. It is a real eye-opener to how well you really have the project times tied down.

What I see in most cases when the view is shifted to the network, it looks like a big ball of string, with no end or no definite beginning. It takes a lot of moving, rearranging, and changing to get it to look like a real actual usable PERT chart.

(Or you can create it from scratch as a network, with a big PROJECT START box on the left and a big PROJECT FINISHED box on the right, and all other task boxes in between and completely interconnected with time constraint lines.)
 
Last edited:
bce123 - yes, I am well aware that not all runs are the same and that size, fittings, bends, height, etc. all come into play. I was hoping that someone might have some good rules-of-thumb for estimating. In short, I am looking for some method to do a sanity check.

Lancie1 - you are correct, I am using the Gantt chart view and, no, I have not switched over either of the other views. That said, though, I have spent considerable time going over each item and linking applicable ones to their appropriate predecessors and assigning start and finish dates where appropriate. It will be interesting to see the other views, though, especially since I have received no training of any kind on the software. I'll let you know.

Steve
 
Lancie1 - you are correct, I am using the Gantt chart view and, no, I have not switched over either of the other views.
I knew it! Can I read minds, or what? I have done enough project schedules to tell when it is a real plan. A group of tasks have to be interlinked together like an electrical network, or there will be loose undefined ends. Task C must be started before Task A, so ON THE NETWORK chart, there should be a constraint line from Task C to Task A. Without that, you are peeing into the wind. Your chart does not reflect the realities of the project, but is just a grocery wish list.

I challenge you just for ONE TIME, click on that menu button and look at the network version of your chart. Only do this if you are prepared to rework it!

Lancie1, no love loss for MS?
I think MS Project is a very good program. The problem is with people trying to use it without a good understanding of what is a network or PERT (Progam Evaluation and Review Technique) chart, and what they are trying to do. If applied correctly, the PERT method will lead to finishing a project exactly on the planned date. Used incorrectly, it leads to project failure.

The real key to making successful PERT (network, not Gnatt) charts is in the first step, drawing a NETWORK chart and assigning each task to a responsible person or foreman AND GETTING THAT PERSON TO AGREE with the time schedule. Once this is done for all tasks, the project success is 95% guaranteed.

Usually Step 1 is skipped, and someone starts making up a Gnatt chart (a list of tasks in some imaginary order that has nothing to do with precedence or time to complete the PREVIOUS task). Of course the actual project work will look nothing like the Gnatt schedule, because it had too many loose ends and did not tie down the start and end dates of EVERY task.

A good PERT chart is arranged as shown, has a Start and End block, has task durations, task start and end dates, and shows the Critical Path, the path of tasks that wil take the longest time to accomplish. This critical path defines the Project Duration, and nothing else does (unlike a Gnatt chart where people type in any old time that they WISH for).

PERT Chart Example.jpg
 
Last edited:
Those who have studied electrical relay logic and PLC programming will notice that the PERT chart looks a lot like a series-parallel logic network. Well, guess what? That is exactly what it is (unlike a Gantt chart which is merely a wish list).
 
great chart lancie 1. really no kidding. but without knowing how long each task will take the only input that steve can enter is the start date and end date of the mech. down time. you have to do a take off of material and from there come up with a labor factor for each item or task to fill in the alloted time for each. as a contractor i estimate jobs all the time the first step is to do the take off. but I really like the pert chart will try it on my next project. Thanks.Steve
 
bce123 - now that you say it, I think I see where my problem is; I haven't (and don't have the background to have) pulled together a realistic take off of materials for the conduit, bends, boxes, etc. for this level of job. I can clearly see what needs to end up where but not what it will take to get it there. If I read you correctly, you need to know this before you can apply your labor factors and accurately allot time.

Lancie1 - I spent some time last night working with my time-line in the network view and, quite frankly, I am not seeing how it will be of much help. I have over 150 items on my list and the network view is huge. It may be that my time-line is badly formatted (groups of logical items like all power conduit runs or all control wire pulls to a specific enclosure are all together and under a title line like "Power Conduit runs to Enclosure A") or something like that.

At this point, I think I need to stick with my Gantt chart view for this project and get some training on MS Project if I'm going to use it differently.

Steve
 
Steve do you have a local contractor that you use on a regular basis? I have a few guys that I run big projects by. They run the data throw their estimating software. The material leadtimes are usually the scariest part that kill timeliness. " I only need 5000' of 1 1/2" IMC" bitten me more than once. I had to buy conduit from the large contractor because they had it at their shop and noone else had it.
 
At this point, I think I need to stick with my Gantt chart view for this project and get some training on MS Project if I'm going to use it differently.
Yes, I think you are right. At least that way you have a time chart that LOOKS like it works.

Remenber, the important thing is not what you put on the chart, it is getting agreement and cooperation from the people who will be responsible for getting the work done.
It may be that my time-line is badly formatted (groups of logical items like all power conduit runs or all control wire pulls to a specific enclosure are all together and under a title line like "Power Conduit runs to Enclosure A") or something like that.
Yes, because every project consists of two types of jobs, those that can be done at the same time (parallel tasks) and those that must be done one after the other (series tasks). Until your task schedule is organized that way, then the actual shortest path to completion CANNOT be calculated, so therefore your Gantt chart does not ACCURATELY tell you the thing that you need to know: How long will it take to complete the job?

Anyone who has done short-circuit and fault-current calculations for an electrical network will quickly realize that a PERT network chart is nothing more than a group of loads (tasks) that create impedance (time duration) in the network, and the path of least resistance (critial path in the PERT chart) is the path the fault current will take.

What does it take to solve an electrical network for the fault current at ANY point? The same as it does to solve the PERT network: A series-parallel diagram of the loads (PERT diagram), the values of all the resistances (task durations), and the power supplies (start date and the workers assigned to each task). If some are missing or not defined, or worse, not connected to the network, then the network cannot be solved. That is the case with 99% of the Gantt charts that are made - they are really useless, unless all you want is a pretty picture.

I have over 150 items on my list and the network view is huge.
There is a PERT method for handling that problem. First, you make a Summary PERT, which simply shows the Major Steps. The Summary PERT is really the only one that the project manager will be interested in at first.

The detail charts are used by the person responsible for that area. Each Major Step then has its own network detail chart, with a Start and End Date. The detail charts start and end dates become the Major Step Start and End dates on the Summary chart.

OT: Back in the old days, we did PERT charts by hand. I still have my PERT slide rule somewhere.
 
Last edited:
Steve,

I feel your pain - been there several times! We've learnt to always add (at least) 10% on to any estimation for BF...(buggeration factor)! It all looks wonderful on a spreadsheet or Gantt chart, but the reality can be very different when you actually get underway šŸ™ƒ(i'm not suggesting you don't know all this - just that sometimes it never gets factored in!).

The other crucial factor is the team you have working for you. In my opinion, it can either make or break the job! If you've worked with them before and maybe built up a good working relationship and understanding - then you're half way there...if it's a relatively new bunch of guys, you may need to allow a little longer. Worst case scenario is they're new to you and will need time to get up to speed (if they are the right calibre)...

Good luck!

Rob
 
steve, there are many programs o there to do estimates here is a low cost one basic but i use it and mdify to fit by creating my own material items IE: plcs,hmi panel building etc.the labor factors are built in for common tasks, i have not found one for what i do. i used to use an excell spreadheet but found it to be to much trrouble to up date and i dont have time to make my own data base from scratch so this works well for me i can update prices via upc codes in excell format wichmy enders send to me daily. if you have a B.O.M. you could just enter it in the take off. and spit out the labor factors. if you dont have one you need one are you will eat up your time buying parts daily and waiting long lead times in the middle of your project. there are even estimating services that can do this for you.but i dont know of one for the controls side. i have built my labor rates from keeping good records of all my controls, panel building. and programing jobs. but this would vary greatly from one person to the next. lets face it some people are alot beter programers or installers than others so labor varies greatly in the automation world. Just my thoughts. with that said if you could tell me how much pipe and wire you have to run and how many I/O you have i could give you a good guestimate, Steve
 

Similar Topics

I have been working for a small turn key machine builder. We mostly do table top machine which do a few automated steps in assembling a product...
Replies
20
Views
5,334
I am doing a revamp of a machine with motion control. I have no data of the inertia of the load on a certain motor, but I should want to know it...
Replies
3
Views
2,110
Has anyone come across any good resources for developing data flow estimates from PLC's, for some given inputs and constraints? We have a network...
Replies
13
Views
4,908
Is there anybody here who is involved with estimating final price for an automation system in a tender (and not necessarilly just in a tender). I...
Replies
10
Views
4,163
Expanding on interesteds topic of How much do companies charge for A programming a new file and B fault finding a customers code. what is the...
Replies
2
Views
3,243
Back
Top Bottom