3.3 How to Read This Handbook · Spatial Intelligence Handbook

Given the interdisciplinary nature of spatial intelligence, which spans AI, computer vision, robotics, graphics, GIS, and cognitive science, it is crucial to provide tailored reading paths that help each type of reader gain the most value from this handbook.

$\blacktriangleright$ For Beginners and Students (e.g., , undergraduates, junior graduate students, newcomers to spatial AI)

•

Begin with Part I: Introduction and Background (Chapters 1–3) to grasp the motivation, historical evolution, and research scope.
•

Then proceed to Part II: Spatial-aware Representation to build a solid understanding of core representations in 2D, 3D, and 4D, including geometry and topology.
•

Optionally skim Part IV: Applications to get a glimpse of real-world use cases, such as robotics, autonomous driving, and AR/XR.

$\blacktriangleright$ For Researchers and Academics (e.g., , PhD students, postdocs, faculty, or interdisciplinary scholars in AI-related fields)

•
Focus on Part III: Spatial Intelligence, from Perception to Generation, which includes chapters on:
- –
  
  Spatial perception, from detection and segmentation to depth and odometry
- –
  
  Symbolic, geometric, neural, and LLM-driven spatial reasoning
- –
  
  Generative spatial models for scenes, videos, and worlds
•

Also study Part IV: Applications for cross-domain synergies, particularly in GeoAI, healthcare, urban simulation, and embodied cognition.
•

Consider Part V: Discussions and Futures for open problems, benchmarking gaps, and research directions.

$\blacktriangleright$ For System Developers and Engineers (e.g., , practitioners building simulation environments, digital twin platforms, or perception pipelines)

•
Prioritize chapters in Part III (Generation) and Part IV (Applications) that introduce:
- –
  
  Modular spatial generation frameworks
- –
  
  Multi-agent simulation engines and control loops
- –
  
  Deployed system domains (e.g., , robotic manipulation, warehouse automation, wearable AI, aerial agents)
•

Use Part V: Discussions and Futures to anticipate where deployment bottlenecks and benchmark gaps lie.

$\blacktriangleright$ For Cross-disciplinary Explorers (e.g., , scholars in architecture, biology, geoscience, or cognitive science)

•

Start with Part I (Introduction and Background) to acquire shared conceptual grounding.
•

Dive into Part IV (Applications) for domain-aligned chapters and illustrative case studies.
•

Refer to Part II (Representation) to gain a technical foothold in spatial structures and encodings.

Overall, this handbook is modular in structure and can be read either linearly or selectively depending on the reader’s focus. The chapter cross-references are designed to aid flexible navigation and long-term use.