TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Launch HN: Bild AI (YC W25) – Understand Construction Blueprints Using AI

98 点作者 rooppal3 个月前
Hi HN! We&#x27;re Roop and Puneet, the founders of Bild AI (<a href="https:&#x2F;&#x2F;www.bild.ai">https:&#x2F;&#x2F;www.bild.ai</a>). Bild AI uses machine learning to read construction blueprints and extract detailed material quantity and cost estimates. Here&#x27;s our demo: <a href="https:&#x2F;&#x2F;youtu.be&#x2F;vBDl7Mh7iEc" rel="nofollow">https:&#x2F;&#x2F;youtu.be&#x2F;vBDl7Mh7iEc</a><p>The problem we&#x27;re tackling is the sheer manual effort that goes into generating material quantities and cost estimates from blueprints today. Contractors and suppliers spend countless hours doing takeoffs by drawing on blueprints by hand - it&#x27;s tedious, error-prone, and costs the global industry $30B a year. A single mistake can lead to thousands in losses on a project.<p>My co-founder Puneet experienced this firsthand as he was building hundreds of houses in Canada. Meanwhile, my background is in applied ML - I started at Google at 19, then Waymo where I built perception models for self-driving cars. Puneet and I met at Hack for Social Impact where we built our first _very_ narrow-scoped prototype.<p>Since then we’ve expanded our scope slowly, with a laser-focus on accuracy. Our approach is to use a suite of specialized machine learning models for specific blueprint comprehension tasks, rather than a single end-to-end model. For example, we&#x27;ve developed computer vision models that are highly accurate at detecting and measuring floor areas, or identifying and counting framing elements like studs and doors. By composing these expert sub-models, we can achieve high accuracy on the overall takeoff.<p>This is somewhat analogous to the approach we took at Waymo for self-driving perception - having an array of dedicated models for tasks like lane detection, traffic light classification etc. It&#x27;s a very different paradigm than the big-data end-to-end models like what Tesla uses - unfortunately we just don’t have enough data yet.<p>We&#x27;re working with some early customers like flooring suppliers to help automate their estimating workflows. But we see a huge opportunity to expand this &quot;AI that understands blueprints&quot; approach across all trades.<p>Would love to get the HN community&#x27;s thoughts and feedback! Construction is an industry I think is really ripe for applying cutting-edge ML techniques. If you have experience in this domain as a builder, architect, estimator, supplier etc. I&#x27;d love to hear about your workflow and pain points.<p>Also if you&#x27;re a researcher or engineer excited about applying state-of-the-art techniques to real-world problems in underserved industries, definitely reach out!<p>We’re currently live with customers but are only able to serve a subset of trades accurately right now. Head to <a href="https:&#x2F;&#x2F;www.bild.ai&#x2F;upload">https:&#x2F;&#x2F;www.bild.ai&#x2F;upload</a> if you want to try uploading a blueprint and we can talk about your use-case. I truly believe we can solve blueprint understanding with AI, and there seem to be a huge number of applications. I&#x27;ll be here all day to chat and answer questions!

16 条评论

BrandiATMuhkuh3 个月前
Congratulations on the launch. This is really cool, and useful.<p>Coincidentally, yesterday I had a client meeting and they ask for exactly that. I&#x27;m working as lead developer for <a href="https:&#x2F;&#x2F;howie.systems" rel="nofollow">https:&#x2F;&#x2F;howie.systems</a> and we are building a co-pilot (knowledge platform) for the AEC industry.<p>Would love to have a talk. Your product could save us lot&#x27;s of work!
评论 #43202173 未加载
djsamseng3 个月前
Awesome! This is a huge opportunity to help a lot of people (clients, subcontractors and builders). A lot of money and time is wasted by the current inefficiencies. We gave takeoff construction plan parsing a go in 2022-2023 but couldn’t get the AI part to work well enough (and still haven’t been able to even with the latest ViT&#x2F; CLIP models). There was a lot of interest though!<p>- You’re right, data is very hard to come by. I’m curious, how do you plan to get around this? Outsourcing human labeling? We found it to be a very difficult task.<p>- The subcontractors and local construction companies we talked to were overwhelming excited about the idea.<p>- It’s entire people’s jobs to get this done and done correctly. They sit on site holding the pdfs in their hands, manually counting and calculating. You bet a lot of mistakes occur. They would absolutely love to have a digital assistant for this.<p>- Some of them (especially managers and owners) are quite technical and are using software such as BlueBeam and other CAD software to make these calculations. It’s quite manual currently, but gives great insight into a better solution. This led us to having the user manually select the symbol they wanted counted (which ML struggled to get right). Just getting the part counts (and highlighting them in the pdf) was a huge help!<p>- Impressive you got square footage calculations correct! In our experience, there was way too much variation between architects (and multistep dimension labeling) which made it hard (even for humans) to get right. How has your model generalized OOD thus far?<p>- Are you planning to integrate voice? Many of the subcontractors we worked with are very low tech. They usually talk with their clients in person, on the phone, or maybe text. But they don’t use email or their smart phones for much.<p>I will be following your work! I have friends who would love to use this once it passes the human threshold.
评论 #43202212 未加载
harmmonica3 个月前
I feel like you must&#x27;ve put this somewhere, but where are you getting the cost information from? If I upload a blueprint and you tell me I need &quot;x 2x4&#x27;s at a cost of $y each totaling $z&quot; for my project where is the y coming from? Can I tell you a specific supplier and then you&#x27;ll &quot;scrape&quot; the cost data? Or maybe I&#x27;m misunderstanding what you&#x27;re doing here.<p>I&#x27;m asking because even though I am (mostly) technically illiterate I have asked both ChatGPT and Claude to help me build a scraper for construction material costs, from the suppliers we use, that can be updated in realtime or at least monthly. Haven&#x27;t done anything with those instructions yet, but I would love nothing more than to use a tool that we could feed a blueprint into and then would tell me, with &quot;laser-focus accuracy&quot; &lt;smile&gt; how many x&#x27;s the project would need and the costs. Even better yet if it could compare costs from suppliers and guide us to the lowest-cost supplier.<p>Edit: oh, while you&#x27;re thinking of replying, how high fidelity do the blueprints need to be? Again, I&#x27;m sure you specify somewhere, but too lazy to find it. How far along the spectrum from &quot;drawn on a napkin&quot; to &quot;fully standardized&quot; do you accept?
评论 #43198294 未加载
serjester3 个月前
I love this, it seems like 90% of the YC AI startups are geared towards selling to devs. There&#x27;s so much value to be had finding where AI is relevant in the broader parts of the economy, especially something like construction.
FloorEgg3 个月前
Curious why you&#x27;re going off of blueprints instead of BIM?<p>The benefit of estimating quantities and cost cycles in with pre-con and business development, the artifacts during the pre-con design phase tend to be different than the takeoff artifacts which are often transformed through BIM.<p>Did you learn something to the contrary? Or are you purposely targeting smaller firms and projects that don&#x27;t use Bim and maybe won&#x27;t for a long time?
评论 #43196663 未加载
评论 #43197292 未加载
评论 #43199333 未加载
评论 #43201611 未加载
nicpottier3 个月前
Who is your market? In my area takeoffs are done by lumber yards, it&#x27;s just a service they provide in order to win the business. So you are going to be selling B2B into a market that is very slow and resistant to change. I imagine many are using some software for their process but I&#x27;ve never seen it.<p>On the other side, architects are using Revit more and more and takeoffs like square footage of flooring are accurate and take no time at all. That&#x27;s another industry slow to change and that used to take more effort so many architects aren&#x27;t providing that information to their clients, but technically there&#x27;s nothing preventing it. There&#x27;s a bit more hand waving when it comes to calculating number of studs etc, but that is pretty straightforward as well.<p>Source: I&#x27;m funemployed as a drafter for a local architect after 25 years in software.
评论 #43196691 未加载
realitysballs3 个月前
It’s interesting space to build a solution. The need for quantity take offs of a 2D PDF drawing that is most likely built in BIM software is mostly driven by contract structure.<p>If owners&#x2F;developers understood this they could create contract structures that incentivize more fluid data collaboration aka the quantity take offs automatically generate as you are designing.<p>Pragmatically though in the current AEC landscape there is still a need for 2D QTO, nice work
ev93 个月前
Seems like an interesting application of AI. Going to be obvious to many, in hindsight. Having run a business in the trades, I used to receive 5-10 emails a day from companies offering estimation services. Clearly there is a need in the market.<p>It looks like your launch is opening this up to the general public - why not niche down to GCs? Maybe the launch is focused on simply gathering more blueprint data to feed your models?
评论 #43196654 未加载
pontifier3 个月前
Very excited for the day that you reverse the process and can generate blueprints and changes on demand.
评论 #43202182 未加载
johanssen3 个月前
Correction: A single estimating mistake can lead to MILLIONS in losses, the bankruptcy of the contractor, and years of litigation... lol!<p>Estimators miss things ALL THE TIME. It&#x27;s the subject of seemingly endless in-house arguments between PMs and Estimators:)
xrendan3 个月前
I love the Blueprint Understanding.<p>One thing I&#x27;ve been thinking about is if you could use a model like this as the first pass for permitters (Like a GitHub Actions CI&#x2F;CD) who review blueprints.<p>Many developers use the regulatory side of various engineering approval processes as a quality control check which costs money and time for the regulator who is tasked with enforcing a standard.<p>It would also be good to speed up the workflow for developers saying hey, this thing looks weird did you really mean to do this?<p>And then further on, you could add a way to check it for constructability. My framer friends often get annoyed at whatever engineer because the way the structure is designed is materially inefficient or hard to construct.
评论 #43199203 未加载
anonu3 个月前
Then you can bid out the materials buying too. This could save homeowners tons of money rather than let a GC manage it.
fzysingularity3 个月前
Congrats on the launch! What kind of computer vision models do you use under the hood?
pkondle3 个月前
Interesting idea. Can i know how accurate or precise is your outcome. From a tech perspective, can you take more about how it works if you can?
bittermandel3 个月前
Congratulations to the launch! Funny how two separate companies launch a product in the same space within a short period of time. AI-BOB, <a href="https:&#x2F;&#x2F;www.aibob.io&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.aibob.io&#x2F;</a>, got a lot of attention during TechArenan in Stockholm last week.
DanKamau3 个月前
Oh hey welcome to the space. I&#x27;m the founder of <a href="https:&#x2F;&#x2F;www.sketchdeck.ai&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.sketchdeck.ai&#x2F;</a>. We started working on the same problem a couple of years ago. Took us longer than I would like to admit.<p>What&#x27;s crazy about this is that the AI revolution is going nuts. We&#x27;ve started with Steel and customers who would traditionally bid on paper are now jumping straight to AI takeoffs. The impact is real.<p>One customer recently told us that he was able to bid on $200M more than he would have been able to otherwise: <a href="https:&#x2F;&#x2F;www.linkedin.com&#x2F;feed&#x2F;update&#x2F;urn:li:activity:7300899005616582656&#x2F;?actorCompanyId=77094417" rel="nofollow">https:&#x2F;&#x2F;www.linkedin.com&#x2F;feed&#x2F;update&#x2F;urn:li:activity:7300899...</a>. That&#x27;s a couple of million in revenue that they would have worked away from because of capacity constraints.