Technical drawing reviewing too time consuming? 15

[Ron88N]

I'm a junior structural engineer with 1.5 years of experience, based in Germany. In my office, we each have to review multiple shop drawings of concrete components every week. The process often feels repetitive and, honestly, a bit demeaning boring, as it mostly involves checking that all components are correctly inserted, referenced, and specified.

Does anyone else find themselves in a similar situation?

Are there any software solutions to help automate or streamline this process? We currently use Bluebeam, but it still feels quite manual.

 

[Ron88N]

But I will continue to use more common definitions of "AI". An early approach was the Turing test, I don't know any structural software that would pass that test.

Definitely, the Turing test is a big challenge indeed. But don't you think that these new AIs (ChatGPT-level) could maybe be able to do part of the drawing review, for example checking that all components are (correctly) inserted, referenced, and specified in the drawing?

 

[DTS419]

This thread reminds why engineers are not typically good businessmen. Why are we paying an engineer's rate to perform quality control checks on a someone else's drawings? Years ago I worked for a precaster who had a guy whose sole job was to check plans. He was not an engineer but had lots of production experience and knew exactly what to look for. Any issues he found he brought to the engineer's attention. This made a lot of sense for everyone involved and was a better more cost-effective solution than the engineers doing it.

On an somewhat related note, the same is true of drafting. Engineers who do their own drafting instead of farming this work to actual drafters are using up time that could bill at a higher rate to perform work that can done more cost effectively and in most cases better by a drafter. There is a saying, "when you let engineers do drafting, you get lesser quality drawings that you paid too much for."

 

[cliff234]

Ron88N - The problem the software doing "some" of the checks is understanding which were done and which were missed; and which were partially done, and which were incorrectly done by the software. In my opinion this will be an accident waiting to happen.

Likewise some constructability problems might be easily solvable if the problematic condition (usually a connection) is properly designed and detailed. But in the U.S., how often do design models contain all the connection details? (Answer: Very rarely)

Engineer testifying in court: "But I thought..."

 

[ThomasH]

Definitely, the Turing test is a big challenge indeed. But don't you think that these new AIs (ChatGPT-level) could maybe be able to do part of the drawing review, for example checking that all components are (correctly) inserted, referenced, and specified in the drawing?

I think that AI may in the future be able to assist with things like that. One challenge will probably be how to describe the task to a computer. I think that is the main thing with AI. To be able to communicate with a computer as you do with a human. That is also what the Turing test is about.

But I think it will be some time before "Checked with AI" will be considered as a guarantee of quality.

 

[XR250]

On an somewhat related note, the same is true of drafting. Engineers who do their own drafting instead of farming this work to actual drafters are using up time that could bill at a higher rate to perform work that can done more cost effectively and in most cases better by a drafter. There is a saying, "when you let engineers do drafting, you get lesser quality drawings that you paid too much for."

As basically a one-man show, I do my own drafting and for the relatively small jobs aI do ($10k USD or less typically), this works better for me and the customer. I have been drafting my work since about 1995 and I can go toe to toe with most dedicated draftsman. I also engineer the plans as I draft them so my "feel" of the job is better than if I delegated it. And, there is no time lost trying to communicate my ideas to the draftsman and then checking their work.

 

[dflva]

OP: Agree, checking shop drawings can be somewhat boring, but it is very important as clearly explained above. In going from school to work you may have been caught off guard by the shop drawing review portion of the job. Not every aspect of working in an office is previewed by a course at engineering school. I expect there was not a class called Shop Drawing Review and now you're asking yourself: "What's this? Where did this come from? Why am I doing this? This is so awful."

Meanwhile, structural analysis/design was emphasized in school and this is where you feel your interest, expertise, and best opportunity to contribute lies - rightly so. Trouble is, you're not in the structural analysis/design business! You're in the Development-Architecture-MEP Engineering-Structural Engineering-Construction-Real Estate Sales Business, of which shop drawing review is a part. After 30 yrs, though less than I once did, I'm still looking at shop drawings.

One more thing, all of the seniors at your office have been through the same thing you're going through - they're observing how you handle it. Don't blow it off - you don't want one of those mistakes, mentioned in other responses, to get by you. Case Studies are full of failures that could have been prevented in the shop drawing stage.

 

[MJB315]

I lost my fingerprints checking shop drawings. I now look back and realize how lucky I was.

My first few years were spent on construction administration for commercial building projects. This was in the '00s, on the cusp of Bluebeam - but not quite.

On one large, very multi-story steel project, each steel piece drawing was printed on ARCH E sheets (36 in x 48 in) - SIZE E! Every W10x12 infill around a floor penetration - 36x48. Three copies. Etc.

Each piece had to be reviewed, and all the comments made on one sheet had to be manually written over to two others copies, each page stamped with a review stamp - and folded (FOLDED) and FedEx'd back to the team. The act of folding several thousand sheets caused my finger prints to, ahem, smooth-i-fy. Which wasn't a good look (nor was it comfortable). Borderline demeaning, even.

But man, did I get a feel for how structural systems come together. And man, did I stumble upon a series of potential errors along the way. Why? I was just another set of eyes on the drawings, double checking, and allowing my intuition (which was still developing) sniff out things that didn't feel right. That helped build that intuition over time. It helped me be able to, eventually, design, depict, and review a structural system.

I mean even this summer, I reviewed shops on a small steel frame that I designed to support a small thing. My drawings were good, but, when I saw how the pieces were coming together from the detailer -- it, just, didn't feel right. Catching it there allowed me to make an easy tweak, which was no big deal, and it came together nicely. Thank you, shop drawing review.

So as an engineer with 1.5 years of experience, one of the best things that you could do is construction administration for projects. That's my two cents. Plus, with Bluebeam, the risk of loosing your finger prints is minimal.

Grow that intuition, where you are, while you can.

-----
Epilogue: Fingers heal and fingerprints restore quite quickly.

 

[Eng16080]

One further point about the importance of the shop drawing/submittal review process in general:

Of all the projects I've worked on which went poorly during construction, the big problems were almost entirely related to aspects of the construction which didn't rely on the shop drawing review process. For this reason, it seems steel building elements rarely have major problems. Field framing, on the other had, often does due to the missing level of oversight.

Shop drawings provide a nice preview of what's about to be done. If some of these carpenters had to create shop drawings for what they were planning to do, we'd definitely be shaking our heads and catching major problems before they happen.