Live Coverage · Google I/O 2026

Google I/O 2026 Keynote
Complete Breakdown

Everything announced — Gemini Spark, Android XR glasses, Antigravity 2.0, new pricing, and a full developer impact analysis.

📍 Shoreline Amphitheatre, Mountain View · May 19, 2026 Gemini Android XR Antigravity Developer

TL;DR — Key Takeaways

• Gemini 3.5 Flash launches today — 4× faster than comparable frontier models, outperforms Gemini 3.1 Pro on coding & agentic benchmarks
• Gemini Spark: a 24/7 personal AI agent that runs in the background, monitoring Gmail, Calendar, and Tasks without opening any app
• AI Ultra subscription cut from $250 → $100/month; daily prompt limits replaced by compute-based metering
• Antigravity 2.0 goes globally available today — the agent-first dev platform now supports subagent orchestration and one-call managed agents
• First Android XR audio glasses confirmed for Fall 2026 — Samsung, Warby Parker, Gentle Monster as hardware partners
• AI Mode in Search surpasses 1 billion monthly active users in just one year
• SynthID has watermarked over 100 billion images and videos

Ten years after Google declared itself an AI-first company, Sundar Pichai opened Google I/O 2026 with a clear message: we are no longer in the era of AI assistants. We are now in the era of AI agents. Every major announcement at this year's keynote — from Gemini Spark to Antigravity 2.0 — was a chapter in the same story: AI that acts independently, not just responds.

Here's every significant announcement from the keynote, organized by theme, followed by a developer-specific impact analysis.

Part 1 — All the Announcements

🤖 New Models

LIVE TODAY

⚡

Gemini 3.5 Flash

General-purpose frontier model. Surpasses Gemini 3.1 Pro on coding, agentic, and multimodal benchmarks. Runs at 4× the output speed of comparable models. Available now via the Gemini app and Gemini API.

LIVE TODAY

🎨

Gemini Omni Flash

Creative-focused model designed to generate anything from any input — video, music, images. Available in the Gemini app, Google Flow, and YouTube Shorts starting today.

🧠 Agents & Platform

BETA NEXT WEEK

✨

Gemini Spark

A persistent AI agent that monitors Gmail, Calendar, and Tasks 24/7 in the background. Demonstrated live drafting emails, preparing morning briefings, and handling multi-step scheduling — without opening the app. US AI Ultra subscribers next week.

LIVE TODAY

🚀

Antigravity 2.0

Google's agent-first development platform. Spin up specialized subagents for complex workflows. Built-in terminal sandboxing, credential masking, and hardened Git policies. Globally available today.

LIVE TODAY

📦

Managed Agents API

A single API call provides a fully provisioned agent with a remote sandbox — no infrastructure setup. Brings the full power of the Antigravity harness via a clean API endpoint.

LIVE TODAY

🛠️

Antigravity SDK

Full programmatic control over the Antigravity agent harness. Deploy customized agents on your own infrastructure. Complements Managed Agents for teams that need self-hosting.

💳 Subscription Overhaul

AI PRO

$20/mo

Existing tier, unchanged. Standard Gemini app access and usage limits.

NEW ↓ 60% OFF

AI ULTRA

$100/mo

5× usage vs Pro · 20 TB storage · YouTube Premium · Gemini Spark beta access (US, next week)

AI ULTRA MAX

$200/mo

Previous $250 top tier, reduced. Same capabilities retained.

⚠️ Important: Google is replacing daily prompt limits with a compute-used model. Complex video or agentic tasks consume more of your monthly budget than simple text queries. When you hit the cap, the system falls back to Gemini 3.5 Flash. Additional credits available on a pay-as-you-go basis.

🔍 Search, Workspace & Apps

1B+ USERS

AI Mode in Search

Crossed 1 billion monthly active users in its first year. AI Overviews now at 2.5 billion MAU. Search is shifting from individual queries to ongoing conversational threads.

LIVE TODAY

Daily Brief

Pulls Gmail, Calendar, and Tasks each morning to generate a prioritized, skimmable digest with suggested next steps. Available today for AI Plus, Pro, and Ultra subscribers in the US.

SUMMER 2026

Docs Live

Converts conversational speech — including false starts and corrections — into a finished Google Doc. Rolling out to AI Pro and Ultra subscribers this summer.

COMING SOON

Ask YouTube

A new conversational search experience inside YouTube. Ask questions about video content without having to watch the full video.

👓 Hardware — Android XR Smart Glasses

FALL 2026

Android XR Audio Glasses

Google's first audio-only smart glasses: camera, microphone, speakers — designed for all-day wear. Compatible with both Android and iPhone. Partners: Samsung, Warby Parker, Gentle Monster. Live demo showed ordering coffee, summarizing messages, and adding calendar events hands-free.

PREVIEWED

In-Lens Display Variant

An optional display version of the glasses is also in development. Google and Samsung showed two designs on stage with more variants expected when the full range launches later this year.

🛡️ AI Safety — SynthID & Content Credentials

SynthID has now watermarked over 100 billion images and videos and 60,000 years worth of audio. Google is adding Content Credentials verification across products — including Circle to Search and right-click in Chrome — so users can see whether content originated from AI or a camera, and whether it's been edited with generative AI tools.

---

Part 2 — Developer Impact Analysis

"The core message of I/O 2026 for developers: you're no longer building features — you're orchestrating agents. The shift isn't cosmetic. It's architectural."

This keynote wasn't really a consumer product showcase dressed up as a developer conference. Nearly every announcement maps directly to a change in how software gets built. Here's a ranked breakdown of what matters most, and why.

Top 5 Developer Announcements — Ranked by Impact

1

Antigravity 2.0 — The Agent Infrastructure Default

This is the most consequential developer announcement of the keynote. Antigravity 2.0 positions itself as the default substrate for agentic applications in the Google ecosystem. Managed Agents removes the biggest friction point — infrastructure setup — by delivering a fully provisioned agent via a single API call. The Antigravity SDK adds self-hosting for teams that need it. Subagent orchestration with built-in credential masking and terminal sandboxing means production-grade security without custom tooling. This is Google's answer to the question: "Where do I run my agents?" — and it's globally available today.

Live Today Managed Agents API Full-stack Impact Self-hosting Option

2

Gemini 3.5 Flash — Frontier Quality at Speed

The model story is simple: frontier-class performance at 4× the output speed, with benchmarks that beat the previous Pro model. For developers, this changes the cost-quality calculus. Running agents on Flash instead of a heavier model is no longer a compromise — it's a sensible default. The Gemini API makes it available today, meaning pipeline updates can start immediately. The compute-based billing model compounds this: faster model + cheaper per-token cost = dramatically lower operational costs for high-volume agentic workloads.

Live via API 4× Speed Gain Cost Efficiency

3

WebMCP — An Open Standard for Browser-Based Agents

WebMCP is the most forward-looking announcement of the keynote. It proposes a web standard that lets developers expose structured tools — JavaScript functions, HTML forms — so browser-based AI agents can execute complex tasks with speed and precision. Chrome 149 origin trial starts now. Think MCP server functionality without a server: the agent lives in the browser and interacts with structured tool definitions exposed directly by the web page. If this becomes a standard, it fundamentally changes how agentic web apps are architected. Worth watching closely even at the experimental stage.

Experimental — Chrome 149 Open Standard MCP in Browser

4

Android Migration Agent — Weeks → Hours

Google previewed an Android Studio agent that automatically migrates React Native, web framework, or iOS code to native Kotlin Android apps. The agent analyzes the source codebase and handles the heavy lifting. Google claims migrations that previously took weeks now take hours. Paired with the open-sourced Android Skills and the stable Android CLI, this signals a broader shift: AI agents aren't just writing new code, they're taking over the unglamorous work of modernization that blocks most teams.

Preview React Native → Kotlin iOS → Kotlin Android Studio

5

Google AI Studio — Full-Stack App Development, End to End

AI Studio now supports native Kotlin for Android vibe coding, Workspace integrations, one-click Cloud Run deployment, and Firebase services — all within the same environment. The ability to export the complete project state to Antigravity closes the loop: prototype in AI Studio, hand off to Antigravity for production. For solo developers and small teams, this compresses the "idea → working app → deployed product" timeline to something genuinely fast.

Live Today Cloud Run Firebase Native Kotlin

Impact vs. Readiness Matrix

Not everything announced is available today or affects every developer equally. Here's how the announcements map across two axes: developer impact (how much it changes what you build or how you build it) and readiness (how soon you can actually use it).

High Impact · Available Now

• → Gemini 3.5 Flash API
• → Antigravity 2.0 + CLI
• → Managed Agents API
• → AI Studio full-stack deployment
• → Antigravity SDK

High Impact · Not Yet Ready

• → WebMCP (Chrome 149 origin trial)
• → Android Migration Agent (preview)
• → Android Halo + Gemini Spark
• → Android XR glasses SDK

Available · Narrower Impact

• → Android CLI (stable)
• → Modern Web Guidance
• → Chrome DevTools for Agents
• → Android Bench leaderboard

Worth Watching · Indirect

• → Gemini Spark (consumer-facing)
• → Android XR audio glasses
• → Daily Brief / Docs Live
• → SynthID expansion

The Bigger Picture

"Google is not building AI features. It's building an AI platform — and this I/O was the clearest statement yet of what that platform looks like."

Three threads connect every major developer announcement at I/O 2026:

1. Infrastructure abstraction. Antigravity 2.0 and Managed Agents follow the same playbook as Firebase and Cloud Run did for earlier generations of apps: take the hard infrastructure problem, abstract it into an API call, and let developers focus on product logic.

2. Agent-native tooling replaces manual tooling. Android Migration Agent, Chrome DevTools for Agents, Modern Web Guidance — these aren't incremental improvements to existing workflows. They're an entirely new category: developer tools designed to be operated by other AI agents, not just humans.

3. The model layer commoditizes further. Gemini 3.5 Flash matching (and beating) the previous Pro model on benchmarks at 4× the speed continues a trend: the frontier model gap closes fast. If Flash-class models are good enough for production agentic workloads, it changes the competitive dynamics for every AI provider.

For architects and BIM developers specifically: The most immediately actionable pieces are the Managed Agents API and AI Studio's Cloud Run deployment. Spinning up a BIM-integrated web tool — say, a Revit parameter checker or a project document Q&A agent — no longer requires standing up backend infrastructure separately. WebMCP is worth a closer look: exposing structured tool definitions directly in browser-based BIM viewers could make AI-driven design review a web standard rather than a bespoke integration.

What to Do This Week

If you're a developer looking to act on today's announcements, here's a concrete starting list:

• Switch any existing Gemini API calls to gemini-3.5-flash and benchmark against your current model
• Explore the Antigravity 2.0 docs and Managed Agents quickstart — the single-API-call agent setup is worth evaluating against your current stack
• Register for the Chrome 149 WebMCP origin trial if you work on web apps with AI integrations
• Check the Android Bench leaderboard to see how different models perform on Android-specific tasks before committing to a model for mobile development
• If you have a React Native app you've been meaning to migrate, put the Android Migration Agent preview on your radar

Bottom line: Google I/O 2026 was the most developer-focused I/O in years — not because of individual feature announcements, but because the entire architecture of the keynote was a statement about how software development itself is changing.

The agentic era isn't a future state. It shipped today. Antigravity 2.0 is live, Gemini 3.5 Flash is in the API, and Managed Agents is one API call away. The question for every development team isn't whether to engage with this shift — it's how fast to move.

All sessions and codelabs from I/O 2026 will be available on demand at io.google starting May 21.

Google I/O 2026 Gemini Antigravity Android XR Developer Tools AI Agents WebMCP Gemini API

AI x Architecture: A 20-Year Veteran’s Journey – Building a “3D Site Analysis System” with Ontology Packs

 

Hello, I’m Atelier Kai, a construction project manager with over 20 years of experience in the field. Throughout my career, I’ve overseen countless complex projects, focusing on quality management and architectural engineering. Today, I am embarking on a new mission: merging human expertise with Artificial Intelligence.

Recently, I’ve been experimenting with the "Architecture Law Ontology Pack" provided by @alex_ai_mcp and participating in the OpenCrab beta testing community. What started as simple curiosity—"What is an ontology in architecture?"—turned into a functional application that I built myself.

Here is a look at my [Integrated 3D Site Analysis System] and how AI is changing the way I work.

---

1. The Power of "Vibe Coding"

Even without being a professional software developer, I was able to translate my decades of industry knowledge into a working application. By accurately describing the "workflow" of a construction professional to the AI, I developed a high-quality HTML-based tool.

This isn't just a pretty interface; it’s built on the logical structure of architectural laws, allowing for a seamless transition from site data entry to 3D visualization.

---

2. Core Features: From Law to Feasibility in 5 Steps

The system is designed to handle the critical initial phases of a project in one go:

• Step 1: Site Data Input: Input address, area, and road conditions.

• Step 2: Automatic Zoning Detection: Identifies the specific zoning district (Residential, Commercial, etc.) and retrieves legal constraints.

• Step 3: Automated Legal Review: Checks compliance for Building Coverage Ratio (BCR), Floor Area Ratio (FAR), setback lines, and solar access.

• Step 4: Feasibility Report (DDFL): Estimates construction costs and sales revenue to calculate projected net profit and ROI.

• Step 5: 3D Mass Modeling: Uses Three.js to instantly visualize the building's volume within legal boundaries.

---

3. Professional 3D Simulation at Your Fingertips

The 3D Modeling section is where the magic happens:

• Layer Control: Toggle visibility for site boundaries, setback zones, parking lines, and the building mass.

• Real-time Adjustments: Adjust sliders for floors or height, and watch the 3D model update instantly to reflect legal compliance.

• Export Plan: Generate and download SVG floor plans based on the analysis for immediate use in presentations.

---

4. Lessons Learned: Data Structure is Everything

This project taught me that while AI is a powerful tool, its intelligence depends on how data is organized. The "Architecture Law Ontology Pack" was the key; it provided the logical framework that allowed the AI to understand the context of complex building codes.

While there are still challenges ahead—such as integrating real-time government APIs for land data—I am amazed that a few prompts could create such a high-quality tool. It has given me the confidence to digitize my 20 years of "offline" know-how into the AI era.

Special thanks to @alex_ai_mcp for the vision and my fellow testers (@sooo.ryu, @kue_0421, @ziteun_vin) for the inspiration.

---

[A Note from Atelier Kai] As someone who values the analog touch of a traveler's notebook and a fountain pen, I find it fascinating to be simulating houses with AI. We shouldn't fear technological change; instead, we should ride the wave. That is where the next great opportunity for architects and builders lies.

Tags: #AIArchitecture #SiteAnalysis #PropTech #ConstructionPM #VibeCoding #Threejs #BuildingCodes #AtelierKai #DigitalTransformation

[Revit] Advanced Door Family: Mastering Visibility with IF/NOR Logic

 

Today’s session at the B[BIM] Revit/Dynamo Self-Study Lab focuses on a cornerstone of library creation: the Door Family. We’re tackling the delicate balance between door frames, panels, and the logic required to toggle between "Open" and "Closed" states without breaking constraints.

---

Problem Diagnosis

The primary challenge in door family creation is the 0-degree constraint failure. When a door panel is closed (0°), Revit often loses the directional reference for the rotation arc, causing the family to "break" when the angle is changed back to a non-zero value. Furthermore, managing visibility—showing hidden swing lines when closed but hiding them when open—requires a rigid logical structure to prevent overlapping graphics or manual errors.

---

Revit-Native Workflow

You can achieve high-performance visibility control within the Family Editor without external tools by using a "Logical Toggle" system.

1. Reference Line Rotation: Never host a door panel directly on a Reference Plane for rotation. Instead, use a Reference Line. Align and lock the end-point of the line to the intersection of the frame and wall faces. Apply the Angle parameter to this line.

2. The Logic Gate (IF/NOR):

   • Create a Yes/No parameter: Is_Open.

   • Create a second Yes/No parameter: Is_Closed.

   • In the Formula column for Is_Closed, enter: not(Is_Open).

   • This ensures a perfect NOR-gate relationship: if one is on, the other is mechanically forced off.

3. Visibility Mapping:

   • Closed State (Hidden Lines): Select the symbolic lines representing the door swing in the closed position. In the Properties palette, link "Visible" to the Is_Closed parameter.

   • Open State (Panel & Swing): Select the open door panel and the 2D swing arc. Link their "Visible" property to Is_Open.

---

Dynamo Design (Batch State Control)

In a large-scale project, manually clicking every door to "Open" for a 3D visualization is inefficient. Use this Dynamo logic to batch-control the visibility states.

Node Sequence:

1. Categories (Doors) > All Elements of Category: Selects all door instances in the project.

2. Element.GetParameterValueByName: Retrieve the current Opening Angle or the Is_Open Boolean.

3. Python Script (Logic Toggle):

   Python

   # Toggle the Boolean state for all selected doors
   input_list = IN[0]
   output = [not x for x in input_list]
   OUT = output
   

4. Element.SetParameterByName: Feed the toggled results back into the Is_Open parameter to update the entire project instantly.

---

Expert Tip

Master the "Sub-Category" for Global Control.

Instead of relying solely on Yes/No parameters, assign your swing lines to a specific Sub-Category (e.g., Plan Swing [Hidden]). This allows you to control the visibility and line weights of all doors globally through Visibility/Graphics (VG) or Object Styles without entering the family editor again.

Workflow Shortcut: When testing constraints in the Family Editor, use the Flex method. Change your parameters in the "Family Types" dialog and hit Apply instead of OK. This keeps the dialog open so you can troubleshoot multiple angles (0°, 45°, 90°) in rapid succession.

 #Revit #BIM #Dynamo #Architecture #FamilyEditor #BIMManager #ConstructionTech

When 'Vibe Coding' Meets the Construction Site: How a Non-Coding PM Builds Automation Tools

 

Introduction: Refining Technology to Reach the Essence – Archive / L'épure

Construction sites are spaces where excessive information and physical noise coexist. Over the past 20 years, managing diverse projects from large-scale civil engineering to high-end residential sites in Jeju, I have always pondered one question: "How can we strip away non-essential administrative tasks and focus solely on the core values of architecture and quality?"

In the past, the answer required mastering complex programming syntax. However, a new wave called 'Vibe Coding' has arrived. It represents an era where domain experts like us—not professional developers—can generate code and automate tasks using only our 'expert intent.' This post explores how PMs with zero coding knowledge can leverage AI to maximize field productivity.

Body 1: From the Era of Syntax to the Era of Intent

If traditional automation was about learning the rules of languages like Python or C#, Vibe Coding is about conveying the user's 'empirical intuition' to AI.

Based on my 20+ years of experience in field management, the best automation tools come not from coding prowess but from a deep understanding of the site's context. Even when building a simple scheduling tool, only an experienced PM can weave in variables like seasonal rainfall or labor supply uncertainties. Now, instead of worrying about syntax errors, we must focus on the essential flow of the task. AI is merely a translator; your decades of field data are what give the tool its life. This is the process of Archive / L'épure—removing technical vanities to leave only the expert's intent.

Body 2: A BIM Coordinator’s Perspective – Practical Applications of Data Refinement

As a BIM CM Coordinator and Professional Construction Engineer, I see three pillars of automation on-site. First is Administrative Automation. Vibe Coding is a powerful weapon for summarizing specifications or drafting proposals for public procurement. Second is BIM Data Integration. Even without deep knowledge of Dynamo or Revit, you can use AI to frame and debug scripts that categorize interference check results, saving over 80% of manual work time. Third is Digitizing Quality Management. Creating logic to detect defects through site photos is no longer exclusive to tech giants. When an expert’s keen eye meets AI’s computational power, safety and quality improve exponentially.

Body 3: Harmonizing Analog Insight with Digital Tools

The ultimate goal of adopting AI and Vibe Coding is, paradoxically, a 'return to a more human life.' I value minimalist record-keeping and analog sensibilities. Technology should be a means to win back time—time to design better spaces, prepare dinner for family, and contemplate 'wellness environments.' Automating inefficient tasks is a form of Digital Detox, clearing out digital noise to secure space in our lives. Even in my 50s, I embrace Vibe Coding because it is the most efficient way to Archive and refine (L'épure) an expert’s noble experience.

Conclusion: The Power of Archiving Changes the Future

The construction industry is conservative, but its data is dynamic. Not knowing how to code is no longer a barrier. In fact, because you aren't bogged down by syntax, you can translate the true 'vibe' of the site into code more effectively. Trust your 10 or 20 years of experience. When you record that experience through the lens of AI, your site will evolve from a simple construction zone into an intelligent space.

---

#Hashtags

#VibeCoding #ConstructionManagement #BIM #Automation #AIinConstruction #ProjectManagement #DigitalTransformation #ArchiveLepure #PropTech #FieldManagement #WellnessArchitecture #RevitAutomation #QualityManagement #NoCode #FutureOfWork

Why Construction Managers Truly Need AI (Gemini, Claude): Value Beyond Report Automation

 

The construction site is an environment defined by chaos, noise, and an infinite number of variables. Having spent over 20 years navigating the complexities of site management, procurement, and asset oversight, I have witnessed many technological shifts. However, the most profound change I see today isn't in the hardware—it is in how we process information.

Currently, many view generative AI like Gemini or Claude as mere tools for "writing reports faster." But as a BIM CM Coordinator who views the site through the lens of data and structural integrity, I believe the true value of AI lies deeper.

It is about distilling the essence (L'épure) from the noise. It is about transforming a chaotic stream of data into a clean, actionable archive.

---

1. Relieving Cognitive Overload: An Archive for Your 'Second Brain'

Twenty years ago, managing a site was about physical presence and memory. Today, the volume of data is staggering. Between thousands of pages of specifications, real-time material procurement logs, and complex BIM models, a manager’s cognitive capacity is often pushed to its limit.

Drawing from my experience in large-scale residential and theme park projects, I’ve realized that the greatest risk isn't a lack of information—it's the inability to find the right information at the right time. AI serves as a "Second Brain" or a digital archive. It isn't just about generating text; it’s about querying a massive knowledge base of project-specific data to extract precise insights instantly.

By integrating AI with knowledge management tools like Obsidian, a manager can move away from "memorization" toward "curation." This allows us to strip away the administrative clutter and focus our energy where it matters most: on-site safety, quality control, and technical excellence.

---

2. Validating Intuition: Turning Field Experience into Quantifiable Strategy

A veteran manager’s "gut feeling"—that nagging sense that a certain schedule is too tight or a specific material won't perform under certain conditions—is an invaluable asset. However, in modern project management, "intuition" alone rarely convinces stakeholders or financial auditors.

Based on my analysis as a BIM Specialist, I’ve found that AI acts as a bridge between professional experience and data-driven proof. By feeding historical site logs and current progress reports into Claude or Gemini, we can simulate risks with a level of granularity that was previously impossible.

For example, when I manage complex licensing or administrative procedures, I use AI to cross-reference past regulatory hurdles with current project timelines. This transforms a "senior manager’s hunch" into a "predictive risk report." It allows us to be proactive rather than reactive, moving from firefighting to strategic governance. This is the manifestation of Archive / L'épure: recording the past to clarify the future.

---

3. Refining Professional Articulation: From Field Notes to Executive Strategy

There is often a "language gap" in construction. The raw, direct language of the field is necessary for getting things done on-site, but it often fails to resonate in the boardroom or with high-end clients. I have seen many brilliant technical managers lose influence because their insights were buried in unrefined reports.

AI acts as a semantic filter (L'épure). It takes the "rough" data from the field—interrupted schedules, material variances, or technical conflicts—and translates them into the language of Value Engineering (VE) and strategic impact.

By leveraging AI, we can produce technical reviews that are not just summaries, but strategic documents that highlight efficiency and cost-saving opportunities. For a manager with a 20-year career, this is how you elevate your personal brand. It’s about being more than a "supervisor"; it’s about being a "technical strategist" who communicates with clarity and authority.

---

Conclusion: The Elegance of Essential Management

Ultimately, adopting AI is not about working more; it is about working with more intent.

Just as the Japandi aesthetic focuses on the beauty of the functional and the minimal, our management style should focus on what is essential. By using AI to archive the vast complexities of a project and distilling them down to their core truths, we achieve a level of professional "L'épure."

Experience is a legacy that AI cannot replace. However, when that 20 years of wisdom is amplified by machine logic, the result is a manager who is not just seasoned, but visionary.

---

FAQ: Implementing AI on the Modern Construction Site

Q1: How can I trust the accuracy of AI-generated technical content? A: AI should be treated as a highly capable assistant, not a final decision-maker. As a professional, your role is to "audit" the AI's output. Your 20 years of experience allow you to spot "hallucinations" or technical errors that the AI might miss. The expert's eye is the final quality gate.

Q2: Is a deep knowledge of BIM required to use AI effectively? A: Not necessarily, but they are highly synergistic. Understanding how data is structured in a BIM environment helps you provide better "prompts" to the AI. Think of BIM as the "source of truth" and AI as the "interpreter" of that truth.

Q3: I’m already overwhelmed with site work. How do I find time to learn these tools? A: Don't view AI as a new subject to study. View it as a way to "talk" to your project documents. Start small: the next time you have to summarize a 50-page 시방서 (specification) or draft a technical query, ask Gemini to create the first draft. The time you save will become the time you use to master the tool.

Technical Guide: Creating a Parametric Table Family in Revit

 

1. Problem Diagnosis: The Importance of Skeleton-First Design

The most common mistake when creating furniture families is modeling the geometry before establishing the constraints. Without a robust Reference Plane (RP) framework, using commands like Blend for tapered legs often results in "Constraints are not satisfied" errors when the table height or width changes. A professional workflow must prioritize the "skeleton" to ensure the geometry follows the data correctly.

---

2. Solution (Revit): Step-by-Step Workflow

• Establish the Framework: Open a "Metric Furniture" template and draw RP (Reference Planes) for the Length, Width, and Height.

• Apply Label Parameters: Add dimensions and convert them into Family Parameters (e.g., Table_Length, Table_Width, Table_Height). Use EQ constraints to keep the table centered.

• Tabletop (Extrusion): Use the Extrusion command to sketch the rectangle. Use the Align (AL) tool to lock the sketch lines to your Reference Planes.

• Tapered Legs (Blend):

  • Use the Blend command for the angled/tapered legs.

  • Edit Base: Sketch the larger circle for the top of the leg.

  • Edit Top: Sketch the smaller circle for the foot of the leg.

• Flexing the Model: Open Family Types and change the values to ensure the geometry adjusts without breaking.

---

3. Dynamo Design: Automating Batch Type Creation

If you need to generate 20+ different sizes (e.g., for a large office project), don't create them manually. Use this Dynamo logic to import data from Excel:

1. File Path + Data.ImportExcel: Load your Excel file containing "Type Name," "Width," and "Length".

2. FamilyType.Create: Generate new types within the family based on the "Type Name" column.

3. FamilyParameter.SetValueByName:

   • Input 1: The newly created Family Types.

   • Input 2: Parameter Name (e.g., "Width").

   • Input 3: Value from the Excel list.

4. Sequence: Excel Data -> List.GetItemAtIndex -> FamilyType.Create -> Element.SetParameterByName.

---

4. Expert Tips for Efficiency

• Naming Conventions: Always use a prefix for your parameters (e.g., ADSK_Width or OFFICE_Length) to keep your data organized when the family is loaded into a large project.

• View Toggle: Use the WT (Window Tile) shortcut to see your Plan, Elevation, and 3D views simultaneously while flexing parameters. This allows you to catch constraint errors instantly.

• Minimalist Aesthetic: When designing for specific aesthetics like Japandi, focus on clean joinery and subtle material parameters to keep the file size lean.

---

Tags: #Revit #BIM #Dynamo #BIMManager #RevitFamily #ParametricDesign #Architecture #VibeCoding #ConstructionTech

Why Is Your Architect Friend Like That? 5 "Incurable Diseases" of the Design-Obsessed

I watched a YouTube video and related to this, so I decided to summarize it.

 

If you have an architect in your life—or heaven forbid, you are dating one—you’ve likely noticed a series of peculiar, often infuriating behaviors. In the media, we are portrayed as romantic, brooding leads. But the reality? It smells like spray adhesive, lukewarm energy drinks, and 3:00 AM desperation.

As a veteran who has spent two decades in this field, I can confirm that these aren't just "quirks." They are professional pathologies developed in the trenches of the design studio. Here are the five most common symptoms of the design-obsessed.

1. The Material Fetish: Wall-Stroking in Public

If your architect friend disappears during a walk, check the nearest facade. We aren't just looking; we are stroking the brickwork or knocking on panels to check for "authenticity." A hollow "thud" from cheap plastic masquerading as stone breaks our hearts.

• The Reality: We are people who realize dreams with other people's money. Since we rarely have the budget to build our own temples, these physical encounters are deeply personal.

2. Funeral Chic: The All-Black Uniform

Gather a group of architects, and it looks like a mid-range funeral. From the "G7" parties at Columbia to local studios, black is the only acceptable color.

• The Hack: It’s not just about "minimalist philosophy." When you’re running on three hours of sleep, "matching" is a luxury. Black always matches black. It’s the uniform of the weary designer.

3. The Mainstream Hater (The Design Hipster)

We have a reflex to criticize anything "Instagrammable." We enter a popular space, cross our arms, and look for flaws to maintain our professional authority.

• The Light-Speed Flip: However, if we find out a space we just trashed was designed by a "Master," our opinion pulls a 180-degree flip: "Actually, the depth of the Master is truly different. You can feel the intention now."

4. The Off-Clock Philosopher: Jargon at the Pub

You want to talk about a funny YouTube video; we want to talk about Peter Zumthor’s concept of "place-ness" while staring hauntingly at the pub's ceiling. Using terms like "spatial violence" in a casual setting is a total mood-killer, but we just can't turn "Architect Mode" off.

5. Whitening Disease (백색증): The War on 4000K+ Lighting

This is perhaps our most clinical-grade hatred. We loathe standard, cool-white overhead lighting. We would literally rather eat our delivery pizza in total darkness—illuminated only by the silhouette of the pepperoni—than turn on a "cheap" ceiling fixture.

• The Struggle: Until we can afford that $3,000 designer floor lamp, we prefer to live like monks in a cave.

---

Conclusion: A Love Letter to the Sincere

While these traits are infuriating, they stem from a genuine desire to live in spaces that actually mean something. We strip away the noise to find the essence—L'épure. It’s exactly why architects so often end up marrying each other; no one else can tolerate a partner who refuses to turn on the lights.

Which of these "diseases" have you caught? Or are you the "Patient Zero" in your friend group?

---

Source & Credit:

• Video: 왜 건축가들은 꼭 ‘이런’ 행동을 할까? (건축가들의 5가지 직업병)