Topic: Draughting Efficiency

[Didge]

I'm curious to see a gathering of views.

How many standard drawing and editting commands would you expect a competent CAD operator to issue in a given 8 hour shift?  By command I'm referring to "Lines" "Circles" "Scale" "Break" "Plot" etc.

I was thinking of around 1250, (ie around 160 per hour).

[Gliderider]

With or without visiting this site a couple of times a day?

[Didge]

   With, obviously 

[M-dub]

With or without visiting this site a couple of times a day?

That is a good question, though, Didge.  I'd be interested to find out.

Sounds like it would be a good challenge, not only to keep a count of this for a user, but to help users increase their efficiency each day.  However, number of commands may not be the greatest measure of efficiency.
Interesting topic, though.

[jonesy]

With or without visiting this site a couple of times a day?

Seriously tho, I dont know whether you could quantify it accurately. Somedays I am always drawing lines and using commands, other days like today I am zoom/pan and checking, and the middle-button zoom and pan doesnt show on the command line, so it wouldnt neccessarily be accurate. Then theres the cries of "can you help me" which could eat into the 160cph

[Bob]

I would imagine that its pretty hard to keep someone going for that long without a break.

Questions arise and the draughtsman has to make decisions or stop the job and ask the engineer.

That decision may based on an internet trawl or digging out other drawings or a site visit or calculations.

Then they are always making cups of coffee, talking about the weekend or retirement

[Didge]

Agreed on all counts guys, I should've indicated a usual lunchbreak.

I got thinking about this during some recent long nights of insomnia and then realised how useful this  information would be for a CAD manager (albeit in a Big Brother Styley I admit)

I agree we spend much of our time chasing drawing originators for further information, drink & ciggie breaks, checking, planning, denying etc,  but in most organisations this time usually gets lost to the same job code as actual draughting. Knowing actual command counts per day would add a cost against these interuptions and also provide CAD managers with a useful tool in identifying staff in need of training or kicking .

I would also be curious to see how salaried staff differ from those staff paid on an hourly basis.

I thought a count of  just the "constructive" commands would weed-out the dialogue surfers, (ie those who spend all day moving dialogue boxes around the screen instead of actually doing anything).

In an organisation of 30 CAD seats, initial and very approximate observations on my part suggest actual efficiency may differ between users by a factor or 20.  A variation of that magnitude seems very difficult for anybody to justify.

Maybe I should just get back on with my work and leave the world to right itself. God only knows what this thread has cost my efficiency.

[Maverick®]

  If one could break those commands into groups tied in with having a certain file open for a reported amount of time it may make the number more usable for what you are trying to accomplish.  Seeing that drawing "a" was open for 2 hours with 200 commands vs. drawing "b" open for 8 hours with 200 commands.  Plus then one could factor in lunch etc. by the time the drawings are open.  There would still be many variables but that would get closer.

  Still...... sounds like allotta work.

[sinc]

I'm curious to see a gathering of views.

How many standard drawing and editting commands would you expect a competent CAD operator to issue in a given 8 hour shift?  By command I'm referring to "Lines" "Circles" "Scale" "Break" "Plot" etc.

I was thinking of around 1250, (ie around 160 per hour).

What exactly are you trying to measure?

Your post title indicates something about "efficiency".  What exactly do you mean?

As an example, I can run Drawing Cleanup, and accomplish in a minute what might take a "competent CAD operator" hours or days to do with commands like "Line", "Circle", "Arc", etc.

So what exactly are you trying to measure?  And what does "number of commands run in an hour" mean?

[Didge]

What exactly are you trying to measure?

Your post title indicates something about "efficiency".  What exactly do you mean?

As an example, I can run Drawing Cleanup, and accomplish in a minute what might take a "competent CAD operator" hours or days to do with commands like "Line", "Circle", "Arc", etc.

Its not so much that I want to measure, it's more a casual curiosity into how best ACTUAL productivity could be calculated. For example, a draughtsman using 1000 standard drawing or editting commands will in principal be more productive than another draughtsman using only 50 such commands in the same time period.  Another example, is it more efficient to employ 3 less experienced draughtsman or to employ 2 experienced draughtsman for the same overall cost.

One approach may be to count total drawings completed, but in most offices many different draughtsman could contribute to any single drawing before it's completion. Measuring time spent on each drawing also seems flawed as many of us work with multiple drawings open while only actually working on one at a time.
Consequently each drawings elapsed TIME would be unrealistically inflated.

I came to the brash conclusion that the only effective measure of a draughtsman's speed was by counting actual commands issued.

I appreciate your comment on drawing cleanup, I would personally count this as one constructive command because I would expect a "competent" CAD operator to be aware of this function.

Hope that makes sense.

[sinc]

I appreciate your comment on drawing cleanup, I would personally count this as one constructive command because I would expect a "competent" CAD operator to be aware of this function.

Hope that makes sense.

Well, not really.

That was just one simple example.  And really, that particular command seems to be very little-known and little-used, despite its incredible power.

Many firms are content to continue doing things in the same-old way.  This leads to "CAD Monkey Syndrome".  The CAD Monkey is an awesomely-fast CAD person - a real wonder to behold.  That CAD Monkey knows how to use command aliases and keyboard shortcuts, and has commands like "TRIM" and "FILLET" down so well that an observer has a hard time telling which commands are being run in which order.  Such a CAD Monkey can build an incredibly-complex layout from Autocad primitives in an extremely short period of time, accomplishing in only a few hours what might take merely a "normal-speed Tech" days to accomplish (assuming "normal-speed Tech" worked in the same basic fashion as "CAD Monkey", just slower).

But a real CAD Tech might be able to come in and do the exact same task with a couple of well-chosen, yet lesser-known commands.  These commands are more complex, and take more thought.  So CAD Tech only runs a few commands over a 30-minute period.  Yet, using only a few commands, CAD Tech takes 30 minutes to accomplish the task.  CAD Monkey is running hundreds of commands per hour, yet takes hours to accomplish the same task.

By your heuristic, the CAD Monkey is the real hot shot.  But I'd rather have the CAD Tech.

I would hazard that counting "commands issued per unit time" is completely worthless.

[Greg B]

To many outside variable would have to be factored in.  Then throw in the stress that goes along with knowing that you are being watch somehow and you can never get a really good feel for how a person operates under normal circumstances.

Counting keystrokes won't help you become more efficient.  Sharing ideas and shortcuts does this.  Make sure the office is open with each other when it comes to drafting and give credit to those that contribute more in time saving, streamlining techniques.

[Didge]

But a real CAD Tech might be able to come in and do the exact same task with a couple of well-chosen, yet lesser-known commands.

Hence the reason for my curiosity, if the benchmark was set against the performance of that "Real CAD Tech" then any command counts below that level must surely be an indication of under-performance (or I accept in very rare cases, the individual may just be an "Exceptional" Real CAD Tech).

If the CAD operators are all using the same software on the same hardware, then in most cases drawings will be created with the similar primitives and modifications. A "Real CAD Tech" may have various shortcuts but in my experience they don't have magic wands (automation being an exception  )

I would propose that in the real world this method of measuring efficiency is probably the best we have,
so I wouldn't agree with it being "worthless" but I do accept it's not a 100% solution.

[Maverick®]

The only way to make that somewhat accurate would be to have all of the people working on exactly the same drawing.

[MickD]

Sinc makes a good point which can be used to your benefit using a 'command logger'. Say you logged the command, time and drawing, you could feed this data into a database and run some queries to find things like certain sequences of commands that appear together regularly and have a look at the drawing and see what was being done. This could be flagged for future automation, even a quick look over the output you could probably find patterns of repeated sequences.
This isn't a big brother tactic either, if uses new this was the reason for using the logger I'm sure they wouldn't mind.