Clash Detection - the pros and cons

[Michael.c]

Seems to me that interference detection is the main component of Revit that Autodesk market their product on.
Yes, of course it has it's uses, but is it all-important?

My experience with it is that we generally don't use it until late in the piece. And when we do, more often than not it comes up with a schedule of thousands of clashes here there and everywhere, which of course, presents us with a very daunting task!

What I strongly believe the tool does, is put the user in a "I won't worry about co-ordinating stuff as I go....I'll just get as much in the model as possible and use clash detection to co-ordinate later" frame of mind! To me this is unacceptable.
Perhaps what is needed is a "live" clash detection system.....one where an alarm bell rings or an error message appears saying "user note: placing duct here will clash with a beam....do you wish to continue" or something along those lines.
One for the wish list, perhaps.

[Steve_Stafford]

It is a major feature of Navisworks Manage that they promote, big price tag for Clash Detective alone. I don't think they put more emphasis on native interference checking, than other aspects of the product (Revit). I'm not convinced that people not using a feature to their advantage throughout the design process is the fault of the feature. Seems like a lack of awareness or rigor on the part of users to me.

It's not like it is particularly hard to use. I wonder if beams are in conflict with pipe? A quick check and I know. It isn't particularly efficient to check every element category against every element category at one time. It can result in a lot of false positives, things that will pass through each other (and be resolved naturally) in a logical construction sequence, like pipe through stud framing and drywall.

Like anything we have to use a tool to our advantage to benefit from having it.

[neightyeight]

There are ways to use the interference report tool while designing. But as Steve said, users have to buy in to the whole idea of coordinating their design to begin with. Usually, I'll lay out some ductwork or pipes, then when that area is complete I'll window select everything I've just made and do a quick interference report with only the elements in my selection against structure or other ducts and pipes. This way the number of "hits" doesn't balloon out of control.

One other note. Many users don't worry about what they don't see. So running ducts or pipes through structure when they can't see it is common. (Bury head in sand...) Doing what I explained above you can interference check selected elements against structure without having to turn stuff on. It helps to keep people's minds on the whole picture, not just their discipline.

[RevitNinja]

Perhaps what is needed is a "live" clash detection system.....one where an alarm bell rings or an error message appears saying "user note: placing duct here will clash with a beam....do you wish to continue" or something along those lines.

People have had a hard enough time getting used to the performance of large MEP models, particularly those in workshared environments. The central file has to manage the permissions of every element in the model and there is more communication happening over the network than most people realize. Can you imagine what would happen if the number of queries increased to the level of being able to "test" every element against the one that you are currently modeling? Even if the feature worked, there is no hardware available on the commercial market that could make it worthwhile. Every time you modeled something, the program would chug, chug, choke, until you gave up and went to get another coffee.

There are certain strategies that can be employed in a model environment to promote three-dimensional coordination. If you get those worked out first and ensure that everyone on the team is willing and able to communicate, then waiting around for a "smart" feature such as this one isn't necessary. From what I've seen (and I've been doing this 10+ years), the mechanical engineer and the electrical engineer in the adjacent cubicle do not talk to each other nearly as much as they should. This is a much, much bigger issue than the clash detection capabilities of software.

[Michael.c]

Thanks for your replys, fellas!

Yes I think you're probably right Revit Ninja, on all counts.

The point that I was really trying to make which hasn't really been commented upon, was the one whereby I feel that what is happenning as a consequence of having this technology, is that the user thinks that he / she can safely ignore the co-ordination issues, happily work away oblivious of all the other services, and use the clash detection tool to fix everything up at the end. Needless to say, there's rarely enough time to do this, and even if there was, it would be the wrong way to work, anyway.

To me, we should co-ordinate as much as possible as we go. ESTABLISH a services zone system and stick with it, and if you can't, TALK to the electrical bloke sitting in the cubicle next to you. RING the architect / structural engineer. TELL the Hydraulic engineer that he has to move his pipe
Yes, there's still going to be clashes, but the intereference check is there exactly as that....a checking tool to pick up the small number that slipped through the net.

[USMCBody]

the auto clash detection like auto spell check would be cool. Just turn a bit of it red. Tell it what items you really care about (beams etc..) But then again I think it would be more of a pain that it is worth. Mostly because of hardware issues, but then there are times I draft things in a way that is not final... Like if I have conduits stubbing into the ceiling I'll draw a conduit around the parameter, make sure the height is above the ceiling, and then tell the outlet I would like to start drafting a conduit and then connect to the loop in the ceiling. Then I'll delete the ceiling loop and done. Sometimes I will have to overdo a turn and then adjust it back to the correct spot just to make it work. Hate for the auto clash detection to be wasting computer resources on a big project thinking what it hits when I'm just going to delete it or move it anyway...

I guess it could work, but I don't have too much faith... Too much to do and too little time for the programmers. It would also mean that every company would have to get 'super computers' for all the drafters rather than have one 'super computer' do the clash detection on a team. Kind of gets expensive and more than likely all of the designers will not have access to the most current model of all the design trades on the job. It would also get very funny if you give your model to another design firm that also does what you do, but just not on this project... A little easier to just give it to the one with the master model (Arch or construction firm) to run the clash detection. Real life can be a B sometimes...

Either way good idea. Definitely something that could be put in so you have the option to...

[phyland]

I believe it is just good practice to review any identified clashes using the in built tools when documenting in revit.
Most projects I have been involved in require someone to attend a (weekly) meeting where collision detection has been carried out across the project team. Normally this will be done via a navis works file provided to the main contractor or architect.
We then receive a collision report from the main contractor or architect that is the obvious starting point for any review of our model.
My point being don't let down your team mates and have them turn up to the meeting with the same clashes week after week.
I think it is the responsibility of the designer to either correct problems that have been agreed to (at the coordination meeting) within the file you have provided or provide an explanation of why the clash cannot be resolved on your own part so as that can be relayed to the entire design team.
I've seen enough models drawn purely in plan without any regard to sectional information, other services, the physicality of the space etc to realise that not everyone is as interested in producing a coordinated model.
As with every tool available in revit it is just that, a tool. What you choose to use it for is up to you.

[USMCBody]

I'll say this concerning weekly design meetings.... Great only if you want to learn the new Revit way. Don't expect to make any money on the project...

A few 'real world problems'

It basically takes a min of a day out of every week. Swapping out a new model, making changes from the last meeting, prepping for the new meeting, the new weekly meeting, etc... So unless your at a design firm that let's you only work on a few projects at a time you'll get behind (I've had 7 projects do on the same day once...) Which would be fine unless I had weekly meetings for all 7 and if I had 7 days in a week I worked... but then again I bet most companies would rather you work on 15 or so projects at once... What if I have to travel for a project? What if I get dumped with a major last min change on any one of those projects??? What if more than one project want's to have their meeting on Wed...

Be prepared to sit there for 2 hours without even getting into your trade.

I say this because I was involved in a weekly min 2 hour project meeting that I was 'ordered' to attend and mostly ran long. 98% of the time they didn't even say my name, company name, trade, or even hay you crazy are you there.... It was nice to understand the other trades a bit better, but then my company is really not paying me to listen to their problems if I really had no say in the outcome. Hay at least it was a webinar.... Other than that, it was a fun time zoning out and waiting for someone to say my name, company name, trade name, hay you, crazy, or some other words of 'endearment' that never came every meeting.... <insert bad meeting remarks here> B-P

I do agree that Revit is just a tool...

[Michael.c]

um....what??
I think you must have put that post in the wrong message board, USMCBody.
I don't see the relevance here.....