And So It Begins

[Yankee Chronicler]

Sometimes I worry we are hoping that the poorer RDPs are just going to magically get better somehow. Change is hard and people need incentive to do it. It's really easy to blame the issues they face in plan review on the building department. In my mind the only way we are going to move forward is if complaints are lodged against RDPs that are committing these cardinal code sins.

I don't think AI plan check will help many of the worse RDPs. The problem is that they take projects for very low fees, with the intention of doing minimal construction documents. That's why it's always a federal case when you deny the permit and ask for revisions to the documents -- because they don't have any money in the fee to do the construction documents right. It's like a religion for some of them. I've wasted more time meeting and arguing with some of them than it would have taken them to just fix the drawings, but they don't want to do that. I think it's because if they admit once that the building department is right, then they don't have any excuse for handing in lousy drawings on the next project, and the next project after that.

 

[arwat23]

I just tried an AI plan check for a small TI project. Not the one Wesley is peddling, but by a ChatGPT model I trained with some California code info so it wouldn't get IBC and CBC mixed up (which happened a lot in the past). Gotta say, I'm kind of impressed...

I tried it on a partial set of plans that we got comments on a while ago. It caught a few of the same comments the plan reviewer had, and caught a few other mistakes we made with our calculations that the plan reviewer didn't mention. However, it's plan analysis ability - actually reviewing the linework on a detail - leaves a lot to be desired, especially with accessibility. It defaulted often to "make sure your plan has [insert requirement]" when it clearly has that shown. Helpful for sure, but not ideal. It also missed some of the more important comments the plan reviewer had. It wasn't terrible - it caught a few missing elements from the plans that we definitely needed to show (again, this was a partial set of plans. Most of those elements were on sheets I didn't have it review). I also needed to ask it multiple questions like "could you review this detail" and "now review this sheet". It seems to skip over some details and tried to focus on specific things. Again, not terrible, but not amazing.

It did a very good job with the life-safety plan through. It gave some admittedly pretty good advice on what could make that plan better, even if the suggestions aren't technically required for this specific project.

It mostly performed as I suspected it would; It was great with written words but struggled with plans. I'll probably use it for a first pass on internal plan checks, but it's requires a lot of oversight. If the program Wesley is trying to sell is based off ChatGPT, it has a solid start. It's still solidly in "tool" territory though. It's nowhere close to being able to plan check by itself.

 

[tmurray]

I don't think AI plan check will help many of the worse RDPs. The problem is that they take projects for very low fees, with the intention of doing minimal construction documents. That's why it's always a federal case when you deny the permit and ask for revisions to the documents -- because they don't have any money in the fee to do the construction documents right. It's like a religion for some of them. I've wasted more time meeting and arguing with some of them than it would have taken them to just fix the drawings, but they don't want to do that. I think it's because if they admit once that the building department is right, then they don't have any excuse for handing in lousy drawings on the next project, and the next project after that.

I completely agree. I don't think AI will fix it. I don't think a building official arguing with them will fix it. I think it is a role for their professional association. I think the issue is that the people who tend to get involved in their professional association tend to be higher performers who are not exposed to the lower performers. Basically, they don't know the problem exists because they would be shocked at how bad some of their fellow colleagues are.

You would be among those that fit into the exception, largely because you've been a building official and have had exposure to a wide array of work from your fellow RDPs.

I like your comments on it being like a religion to some. I certainly have heard some proselytize to me over the years on how they need to do things the absolute cheapest way for their customer. Even at the expense of regulatory requirements around submissions, sometimes even at the expense of technical requirements they don't believe should apply within a given set of circumstances.

Like all problems, This is a multi-dimensional issue that needs a relatively complex solution. Unfortunately, our brains seem to like to oversimply problems and look for a single solution. The silver bullet that at once solves every issue. I think we need to improve the lowest performing RPDs before we can really see the advantages of AI (at least for building departments).

 

[arwat23]

I like your comments on it being like a religion to some. I certainly have heard some proselytize to me over the years on how they need to do things the absolute cheapest way for their customer. Even at the expense of regulatory requirements around submissions, sometimes even at the expense of technical requirements they don't believe should apply within a given set of circumstances.

I'm not going to defend the way some RDP act, but in my area this is the norm. In our case, we need to compete with the professionals who will be cheap just to get a job, or those who offshore their work to some drafting farm out of south-east Asia for cheap.

I like to think our firm does a pretty good job with the projects we work on, but we lose a ton of projects because our fees are realistic. There have been more than a few times the firm I'm currently at has almost gone out of business due to lack of work, losing relatively huge projects due to some other architect spitting out a much lower number than what's even feasible. Our costs, btw, are often less then what the client ends up spending on an architect that low-balls the client and then blames the city for everything. A few years ago we lost a medium-size TI project because the tenant thought our fees were "laughable". About a year later, we heard that the tenant abandoned the project after spending considerably more than what our proposal was because the architect "didn't know the city's requirements" (the city has no special requirements - typical TI plans would have been sufficient).

At some point, to survive, you need to lower yourself to their level, at least a little bit. Quality will only get you so far when money is involved, especially when a client doesn't know much about the field. Again, not defending how some RDPs act. Just explaining my perspective on the matter.

 

[socaldano]

After several conferences with visits to AI Plan Check companies at their booths, the official marketing has begun.

https://plancheckpro.ai/about-us/

I have soooooo many questions. The last time I spoke, the different vendors said that their system was limited in scope and still in the testing and design phase, but marketing was on full blast. This is a conversation we must have.

Does anyone have any direct experience with something like this? More to come.

I can read so many questions into what you ask.
As an ABET Accredited Computer Engineer, with a duel degree in comp eng and comp sci...
First thought comes to mind "garbage in/garbage out". I would use the AI tool as a sanity check, but use your professional expertise and research to validate whatever AI program tells you. There are lots of stories of test pilots who died because of computer errors, and test pilots who lived because they outthought the computer, which was screwing up.

That being said, ChatAI does a wonderful job on saving me many hours of editing, and keeps me from getting caught in that vicious editing circle. Gives me great case law and other information I did not previously know.

What is scaring me, is the difference between "The Science" proper noun and the science is simple. "The Science" someone is in charge, someone edits, someone controls. The science has young whippersnappers coming up with insane claims like the law of physics is wrong, it fails approaching the speed of light (einstien, E=MC squared), that light actually has mass and the theory of quantum physics. In our world some STUDENT informed one of the greatest architects in the world that his amazing building in New York is going to blow over... and damn if the student wasn't right.

Also, Remember, you can never Trust the Vendor... you must verify everything they promise...and they can break things in the next release. We held off putting in security updates and other updates from EVERY software and OS vendor for six months unless we HAD to put it in. Because we were going to let someone else fall. Anyone remember when the Pentium chip was screwing up on the rounding errors?

I was a specialist at "automation" which I combined AI training with other systems programing skills to automate complex processes. Oversimplified it is mostly creating and updating lookup tables, an expert making decisions and the coder putting it into practice (I was usually both).

Bad people or unqualified people with access to the AI systems can do incredibly bad things. I can tell war stories of what unqualified or stupid people did to good code. I don't know how complete the AI can be on your local systems, it may need to crawl the web to get its information...But the software company should be able to create dictionaries and look up tables and the like for "air gapped" AI systems for building codes.

I know how to feed bad information that will do horrible things. An evil project in my head could do terrible things to the DNA databases.... simply paying families for thier DNA swabs, then submitting them under the data of other families... How are you going to fix it? Another evil project in my head was how easy it would be to irreparably destroy a document database.... simply swapping file names on files and overwriting files, maintaining the same file dates, do 100 a day, when they discover it a couple of years later...how do you recover?
So, one of the things to HELP keep bad things from happening in codes from AI, is SOC II audits of the AI software.

P.S. I would have run this through ChatGPT for clarity and better grammer but the admin indicated you don't want AI editing done on this sight

 

[socaldano]

I'm not going to defend the way some RDP act, but in my area this is the norm. In our case, we need to compete with the professionals who will be cheap just to get a job, or those who offshore their work to some drafting farm out of south-east Asia for cheap.

I like to think our firm does a pretty good job with the projects we work on, but we lose a ton of projects because our fees are realistic. There have been more than a few times the firm I'm currently at has almost gone out of business due to lack of work, losing relatively huge projects due to some other architect spitting out a much lower number than what's even feasible. Our costs, btw, are often less then what the client ends up spending on an architect that low-balls the client and then blames the city for everything. A few years ago we lost a medium-size TI project because the tenant thought our fees were "laughable". About a year later, we heard that the tenant abandoned the project after spending considerably more than what our proposal was because the architect "didn't know the city's requirements" (the city has no special requirements - typical TI plans would have been sufficient).

At some point, to survive, you need to lower yourself to their level, at least a little bit. Quality will only get you so far when money is involved, especially when a client doesn't know much about the field. Again, not defending how some RDPs act. Just explaining my perspective on the matter.

Might I ask what is the difference between a registered design professional (RDP) and a licensed Engineer or Architect? Are they the same, because Missouri statute 327 identifies Engineers and Architects... The other is the TI, which I assume is anything that needs sealed plans which is NOT structural. if sealed plans are required but not structural its TI by an RDP? Am I getting this right?

 

[arwat23]

Might I ask what is the difference between a registered design professional (RDP) and a licensed Engineer or Architect? Are they the same, because Missouri statute 327 identifies Engineers and Architects... The other is the TI, which I assume is anything that needs sealed plans which is NOT structural. if sealed plans are required but not structural its TI by an RDP? Am I getting this right?

My understanding is they're the same thing. Or at least the way I used it in my previous comment (RDP = architect).

TIs (tenant improvements) are just improvements to a tenant's space. Most TI projects we work on would need to be signed and sealed by an architect and/or a structural engineer. These projects go beyond what's allowed to be designed by a non-licensed person in our state.

Not all TI projects require a signed and sealed plan. Here's a link to a very simplified version of what each licensed design professional can and can't do, as well as what unlicensed persons can and can't do: https://ccidc.org/wp-content/uploads/2020/09/DESIGN-LIMITATIONS.pdf

 

[Plumb-bob]

Well, I am looking at alternative career fields with the understanding that AI may replace us eaters. And I have said too many bad things about computers and the internet, so they will come for me first.

Kinda kidding...haha....

 

[Yankee Chronicler]

I completely agree. I don't think AI will fix it. I don't think a building official arguing with them will fix it. I think it is a role for their professional association. I think the issue is that the people who tend to get involved in their professional association tend to be higher performers who are not exposed to the lower performers. Basically, they don't know the problem exists because they would be shocked at how bad some of their fellow colleagues are.

The professional societies probably can't help. There is no requirement for an architect to belong to the American Institute of Architects, and certainly the unlicensed frauds who like to call themselves "architects" even though they aren't can't join the AIA, and wouldn't magically start doing things right even if they could join.

You would be among those that fit into the exception, largely because you've been a building official and have had exposure to a wide array of work from your fellow RDPs.

I was an architect and trying to do it right for twenty years before I got my license as a building official.

I like your comments on it being like a religion to some. I certainly have heard some proselytize to me over the years on how they need to do things the absolute cheapest way for their customer. Even at the expense of regulatory requirements around submissions, sometimes even at the expense of technical requirements they don't believe should apply within a given set of circumstances.

I had a boss once respond when I pointed out a code issue, "I've never understood that section of the code, so I don't think it applies to me."

 

[bill1952]

I'm for a return to drawing board, mylar, slide rule, hard copy codes, and a diazo machine. I love that smell of ammonia.

 

[Yankee Chronicler]

I'm for a return to drawing board, mylar, slide rule, hard copy codes, and a diazo machine. I love that smell of ammonia.

Mylar ?!?!

What about linen? There's nothing like India ink on linen ...

 

[Yankee Chronicler]

Update (of sorts) regarding my trial of PlanCheckAI -- it hasn't happened.

I clicked the provided link, expecting that I would be able to access the site and start using the software. I would not have been surprised if I was asked to create a user name and password. But that's not what I experienced.

Instead, I received an e-mail from Wesley, sent at 3:18 p.m. on Monday, asking if I would be available for a 15 minute meeting at 3:00 p.m. I responded that that was impossible, and -- by e-mail exchange -- we set an appointment for 2:00 p.m. on Thursday.

Confession: Wesley's e-mail said "meeting," and my brain read "telephone call." Thus, I was surprised when today's e-mail brought at invitation (sent at 11:19 p.m.) to a Google Meets.

I suppose this is what the world is coming to. Everybody wants to use Zoom or Teams (or, apparently Google Meets). I've never used Google Meets. In fact, I didn't know it existed until I received this invitation. I doubt I'm set up to do it -- I don't have a web cam or a microphone on my computer. There is a telephone number through which I can call in, but ... WHY?

I just want to try the software. Color me old fashioned, but IMHO if you post a link to a trial version of some new software, that link should allow anyone who clicks it to immediately access the software and try it. At this point, I still haven't seen the software and I am already massively disinclined to be interested in it.

 

[Yikes]

I'm for a return to drawing board, mylar, slide rule, hard copy codes, and a diazo machine. I love that smell of ammonia.

Back then you could tell how bad or unproductive a person's day was by looking at the amount of eraser dust on the carpet under their drafting table.

 

[Yankee Chronicler]

Back then you could tell how bad or unproductive a person's day was by looking at the amount of eraser dust on the carpet under their drafting table.

 

[Joe.B]

In our world some STUDENT informed one of the greatest architects in the world that his amazing building in New York is going to blow over... and damn if the student wasn't right.

Is this a metaphor or a reference to an actual event? If the latter, please elaborate.

 

[arwat23]

I suppose this is what the world is coming to. Everybody wants to use Zoom or Teams (or, apparently Google Meets). I've never used Google Meets. In fact, I didn't know it existed until I received this invitation. I doubt I'm set up to do it -- I don't have a web cam or a microphone on my computer. There is a telephone number through which I can call in, but ... WHY?

Because they (probably) want to show it off in a controlled setting, something they know will work they way they're expecting it to work. At least, that's what I would do, and that wouldn't work on a phone call. Or maybe they're trying to curate who uses it.

I got sent the same email and I can't do the time they asked for. Still waiting on a response for a new time.

I just want to try the software. Color me old fashioned, but IMHO if you post a link to a trial version of some new software, that link should allow anyone who clicks it to immediately access the software and try it. At this point, I still haven't seen the software and I am already massively disinclined to be interested in it.

It makes me wonder why it isn't just available. They can take precautions so people can't abuse it or use it more than x number of times. It's not that hard to set up.

Keep in mind this is an engineering firm that released this program. They could very well be using data from the plans they get sent to optimize their own work. Just speculation based on what could be happening. I got nothing to back it up. But I wouldn't be sending them anything they could use if I was competing with them.

 

[Yankee Chronicler]

I'm not competing with them. I have several projects on which I have done plan reviews that generated significant numbers of comments. I would very much like to feed these into their program and see if they catch everything I caught, and also to see what they catch that I missed. However, life's too short to play telephone tag games. Send me the URL and the activation key, and let me try it. If it's not transparent enough that I can log in and use it -- it's not ready to be deployed. If it reaches the stage of commercial release, I very much doubt they'll convene a Google Meets tutorial with each and every new user.

And, at this stage, there shouldn't be a "use it more than x number of times" factor. Wesley posted here that this current trial isn't limited, that we can use it as much as we want.

 

[Inspector Gadget]

I while ago, I had an appeal where the RPD was mixing Part 9 and Part 3 assemblies. I really struggled with if I should report this person to the engineering society. He was claiming to be skilled in fire protection engineering. My worry was that this would cause people to be too cautious in accessing the appeal system, thus denying them access to justice. I ultimately decided that was not my role, but I always hoped the originating building department had filed a complaint against him.

I have yet to report an engineer or architect to their respective societies. I suspect if I did, various people - not the least of which is my employer - would have issues with that.

Part of the issue, I think, is that engineers aren't really that concerned about fire protection and part 9/part 3 assemblies. I sure as shooting am.

I'm also getting old, and growing to the point where I just don't GAF about the response for doing the right thing.

the issues they face in plan review on the building department. In my mind the only way we are going to move forward is if complaints are lodged against RDPs that are committing these cardinal code sins.

Translation: want better product? Be a better professional.

Update (of sorts) regarding my trial of PlanCheckAI -- it hasn't happened.

I clicked the provided link, expecting that I would be able to access the site and start using the software. I would not have been surprised if I was asked to create a user name and password. But that's not what I experienced.

You're getting the sale treatment. I have only my gut and some basic understanding of AI to work with but believe me, if the samples I've seen are anything to go by, we are years -literal years -from AI being able to plan check.

Lemme tell you what this scam artist really wants: access to submitted, copyrighted plans so that his AI can train on those copyrighted works.

That's what this is all about.

AI is inherently theft.

 

[socaldano]

Is this a metaphor or a reference to an actual event? If the latter, please elaborate.

https://onlineethics.org/cases/will...ine-story-crisis-lesson-professional-behavior lots of other references.

 

[socaldano]

I have yet to report an engineer or architect to their respective societies. I suspect if I did, various people - not the least of which is my employer - would have issues with that.

Part of the issue, I think, is that engineers aren't really that concerned about fire protection and part 9/part 3 assemblies. I sure as shooting am.

I'm also getting old, and growing to the point where I just don't GAF about the response for doing the right thing.




Translation: want better product? Be a better professional.



You're getting the sale treatment. I have only my gut and some basic understanding of AI to work with but believe me, if the samples I've seen are anything to go by, we are years -literal years -from AI being able to plan check.

Lemme tell you what this scam artist really wants: access to submitted, copyrighted plans so that his AI can train on those copyrighted works.

That's what this is all about.

AI is inherently theft.

Your assessment is dead on... The trade for using AI, is their knowledge DB grows.. and if they are stealing professional work product, this is the easy way to get many to make it public.

 

[socaldano]

My understanding is they're the same thing. Or at least the way I used it in my previous comment (RDP = architect).

TIs (tenant improvements) are just improvements to a tenant's space. Most TI projects we work on would need to be signed and sealed by an architect and/or a structural engineer. These projects go beyond what's allowed to be designed by a non-licensed person in our state.

Not all TI projects require a signed and sealed plan. Here's a link to a very simplified version of what each licensed design professional can and can't do, as well as what unlicensed persons can and can't do: https://ccidc.org/wp-content/uploads/2020/09/DESIGN-LIMITATIONS.pdf

Thank you for that...
and that pamphlet also explained why the local building inspector put impossible testing requirements on my wanting to build a small steel framed patio frame and cover. As a skilled and properly trained in arc welding, I wanted to do it in square tube 1/4" wall mild steel. Suddenly the plans went from any engineer or architect to a structural engineer both sealed plans and inspection... AND a firm who would xray and certify each joint..

 

[Yankee Chronicler]

I have yet to report an engineer or architect to their respective societies. I suspect if I did, various people - not the least of which is my employer - would have issues with that.

I guess every state is different, and I don't know what the legal/licensing situation is in Canada. In the United States, professional societies are not the licensing bodies and they have no power or authority to regulate behavior or practices. It would accomplish nothing to report a bad architect to the AIA, or a bad engineer to CASE or the NSPE. Those are voluntary membership organizations. They have zero regulatory authority.

In my state, architects, engineers, land surveyors and landscape architects are licensed by the Department of Consumer Protection. That's who bad professionals should be reported to. However, you are correct. At my former position, my boss and the administration had no objections to reporting architects who consistently screwed up, and/or people who practiced architecture illegally without being licensed.

At the recent job (which was eliminated as of July 1 due to department budget cuts), the boss supported reporting these clowns, but the administration made it clear that they didn't want us to do it. I am quite certain that's because at least two of the major offenders/frequent flyers are politically connected.

If you're going to report someone -- report them to the correct place: the office in your state government that issues the licenses. NOT the professional societies.

 

[Inspector Gadget]

If you're going to report someone -- report them to the correct place: the office in your state government that issues the licenses. NOT the professional societies.

Canada be different.

 

[tmurray]

I have yet to report an engineer or architect to their respective societies. I suspect if I did, various people - not the least of which is my employer - would have issues with that.

The Office of the Fire Marshal reports the most RDPs. Technical Inspection Services is number two.

We feel it is our professional obligation to notify the society when we see professionals who are clearly practicing beyond their capabilities. In cases our staff have been involved with, professionals have been censored and some have had their ability to practice revoked. I don't feel bad. If you can't perform design to code, you need to find another job. I'm not going to bury innocent people to spare someone's feeling.

The only time my employer was made aware of a complaint I made (back when I was working for a municipality) was the time an RDP threatened to go to council to get me fired and they only found out because I told my leadership (leadership decided to send a letter saying the treatment of their employee was unacceptable and if they did not correct their attitude municipal staff would not have to work with them and it would be unlikely they would be able to effectively practice in that jurisdiction as a result). The results of the complaint were that the society said if it happened again, the professional's ability to practice would be revoked entirely.

Your employer should get no say as to whether you report a professional or not. You know the gravity of the violation. They don't. They can clutch their pearls all they want. If they don't get reported, they will never get better.