Glossary
BIM (Building Information Modeling)
A digital representation of a building's physical and functional characteristics. Foundation for many AI applications in the built environment.
Computer Vision
AI that interprets visual information. Used in construction for safety monitoring, progress tracking, and quality control.
CRAFT Framework
A prompt engineering approach: Context, Role, Action, Format, Tone. Helps produce more relevant AI output.
Digital Twin
A virtual replica of a physical building or infrastructure asset, continuously updated with real-world data via IoT sensors.
Generative AI
AI that creates new content — text, images, designs, code — rather than just analyzing existing data.
Generative Design
AI-powered design exploration where the algorithm generates many design alternatives based on specified constraints and goals.
Hallucination
When an AI generates plausible-sounding but false information. Always verify AI output against authoritative sources.
IoT (Internet of Things)
Networked sensors and devices embedded in buildings and equipment. Provides data that powers AI analytics and digital twins.
LLM (Large Language Model)
A type of AI trained on vast text datasets, capable of understanding and generating human-like text. Examples: GPT-4, Claude, Gemini.
Machine Learning
A branch of AI where systems learn patterns from data without explicit programming. Used for prediction, classification, and optimization.
MEP (Mechanical, Electrical, Plumbing)
Engineering disciplines for building systems. AI helps optimize system design, energy efficiency, and coordination.
Natural Language Processing (NLP)
AI that understands and generates human language. Used for contract analysis, specification drafting, and document search.
Parametric Design
Design methodology where parameters control geometry, enabling rapid exploration of variations. Often combined with AI for optimization.
Predictive Maintenance
Using AI to predict when equipment or building systems will fail, enabling repair before breakdown.
Prompt Engineering
The practice of crafting effective instructions to AI systems to get better, more consistent results.
RAG (Retrieval-Augmented Generation)
AI technique that combines text generation with retrieval from specific knowledge bases (codes, specs, manuals) for more accurate domain-specific answers.
Sustainability Modeling
Using AI to simulate and optimize a building's environmental impact across materials, energy use, and lifecycle.
Synthetic Data
Artificially generated data used to train AI models when real data is scarce or sensitive. Growing use in construction safety applications.
Tokens
The units of text AI models process. Roughly 0.75 tokens per English word. Most AI billing is based on token usage.
Training Data
The dataset used to teach an AI model. Quality and diversity of training data fundamentally determine model capability.