Summary

Orbit is a single place to save and re-find online materials—links, PDFs, screenshots, notes—and turn them into understanding. The goal is not “better storage”, but sense-making: helping people remember why something was saved, see how items connect, and share a meaningful path with others.

Research materials often live across many tools (docs, web links, PDFs, images, notes), and the real work starts later: connecting the pieces into an argument. Current tools are good at collecting, but weak at building a clear “line of thinking”, especially in collaboration.

Orbit replaces folder logic with a visual space that supports three simple actions: Collect → Filter → Compose. Users collect items from the web, filter them quickly using keywords and lightweight controls, then compose selected items into Trails—shareable sequences that communicate reasoning instead of dumping a chaotic pile of files.

In short: Orbit is not “more folders”. It is a canvas for collaboration over saved items.

Desk Research

Traditional file management still depends on folders, hierarchies, and naming conventions. That can work for clean, local files—but it breaks down once research becomes cloud-based, link-heavy, and spread across tools.

Main problems observed in research workflows:

• Context gets lost: later it’s hard to remember where an item came from or why it mattered.
• Connections are invisible: saved items stay as isolated fragments, not part of an evolving understanding.
• Collaboration becomes messy: sharing often happens in chats as “20 links”, without structure or reasoning.
• Retrieval costs time: the most painful moment is needing one specific item and not remembering where it was saved.

This project focuses on an alternative mental model: instead of organizing files into containers, support finding, filtering, and explaining a research process.

Use Case — Sara (Design Researcher)

Sara is a design researcher studying how people personalize and troubleshoot smart home ecosystems. During research she saves hundreds of items: forum posts, PDFs, screenshots, product documentation, and notes. Later she struggles to compare them, re-find them, and explain progress to others.

Step-by-step story

1. While browsing, Sara finds a useful forum thread about a smart home automation bug.
2. She saves it into Orbit (link / screenshot / PDF—whatever the material is).
3. Orbit immediately places it into her visual space, without asking her to build folders first.
4. Later, when she needs sources for writing, she opens Orbit and uses filtering (keywords + quick controls) to narrow down fast.
5. Sara selects the most relevant items and drags them into a Trail—a structured path that tells the story of the topic.
6. She shares the Trail (not her whole archive). Colleagues can add one missing link or a stronger source directly into the Trail.
7. Sara returns to writing with a cleaner, explainable set of sources that represents her reasoning.

Our approach

Orbit is built around three actions:

1) Collect

Save any online material into one place: links, PDFs, screenshots, images, notes. The system avoids forcing early organization while the user is still exploring.

2) Filter

Instead of deep folder navigation, Orbit supports fast retrieval through:

• keyword-based filtering (keywords extracted from notes / saved context)
• lightweight view filters (e.g., clusters, timeline/spiral, trails)

3) Compose (Trails)

The key feature is Trails: users drag selected items into a sequence that represents a line of thinking. Trails can be shared, discussed, and extended by collaborators—so sharing becomes “here is my reasoning” instead of “here are random files.

Final draft

Orbit Space (Explore)

A zoomable visual space that holds all saved items. Items can be rearranged by different modes:

• Clustering: related items appear closer based on metadata/notes
• Timeline / Spiral: browse items chronologically
• Trails overlay: show connections created through Trails

Keywords (Context to re-find things)

Orbit surfaces keywords in the center (extracted from notes made while saving). Clicking a keyword filters the space instantly. A “refresh” action can surface new keywords to encourage exploration.

Trails (Explain + Share)

Trails are curated paths made of saved items. They can be set public/shared, and collaborators can contribute directly to the trail without seeing the entire private archive.

Procedure

The project started from the observation that online research materials are fragmented across tools and hard to re-find later. Early iterations explored visual organization and clustering. Feedback pushed the project toward more concrete workflows: collecting from the web, filtering quickly, and making collaboration meaningful.

The final concept shifted from “a cooler file view” to a workflow-focused system: Collect → Filter → Compose (Trails), designed to support sense-making and collaborative progress.

Reflection

What worked well:

• Trails make collaboration clearer by sharing reasoning, not just data dumps.
• The system reduces the pain of “I saved it somewhere but can’t find it again.

What was difficult:

• Designing a visual system that still stays practical and fast for real research work.
• Balancing “exploration” vs. “structure” without turning Orbit into another folder system.

If extended further:

• More robust “relevance signals” beyond tags (usage, recency, citations, collaborator activity).
• Stronger collaboration feedback loops (notifications, additions, trail updates).

Sources

• Vannevar Bush — As We May Think (Memex, trails as a mental model)
• Alexander Obenauer — Lab Notes (modular systems and linked materials)
• Course discussions and feedback sessions (Prof. Boris Müller)
• Related references mentioned during development (e.g., inspiration from web-based collecting tools)