Networked Expandable Terminal
A portable workspace that secures tools in microgravity and keeps unfinished work readable to whoever comes back to it.
AI video demonstration generated with Google Gemini, edited by Jimmy Huang.
It depicts NET’s intended behavior, starting from reliable restraint in microgravity.
Context
University of Washington MHCI+D capstone
Spring–Summer 2026
Team
My role
Led interaction design and systems framing; contributed to research, physical prototyping, and visual storytelling.
Space industry advisor
Sana Sharma · Aurelia Institute
Research question
How can shared workspaces in orbital habitats keep their meaning across interruptions, handoffs and absent colleagues?
Recognition
Presented at SpaceCHI 2026,
NASA Ames Research Center
Project status
Research concept and scenario-based interaction exploration · Seeking sponsors for fabrication and evaluation
Product overview
NET keeps tools connected to the work they support
A repair pauses. Its tools remain secured nearby. Another crew member needs the wrench, but is it still needed for the repair, and where should it go afterward?
Networked Expandable Terminal (NET) is a portable workspace concept for temporarily securing and organizing items in microgravity. It brings together adaptive restraint, expandable containment, and information about the activity those items support.
The concept follows a lifecycle: establish, evolve (use, pause, share), and reconcile. It helps people distinguish what is still needed, what is available to others, and what must be returned as work changes.
Our capstone developed physical explorations and a system concept. The maintenance scenario below illustrates its intended behavior; the full sensing, adaptive material, and inventory integration remain to be implemented and evaluated.
Adaptive restraint and expandable containment
A conforming textile surface accommodates different object shapes
The device expands on demand into a contained volume while keeping contents visible
Keeping object movement understandable
NET distinguishes accounted-for movement from changes needing attention: the light strips glow in the state's color, and the dial at one corner shows its shape. Evolve shows two moments, an item used elsewhere and an unaccounted change.
Context & desk research
Stowage takes space and crew attention
Our industry advisor proposed exploring the transition toward commercial habitats in low Earth orbit. NASA plans to support commercial successors to the ISS, and its transition report anticipates government crews, commercial providers’ astronauts, visiting researchers, and tourists. This prompted us to consider how people with different roles, lengths of stay, and familiarity with a habitat might organize and share its contents. [1, 2]
ISS interior — in use. Image: NASA.
Commercial habitat concept — proposed interior. Image: Axiom Space.
We explored life and work in orbit through astronaut accounts, habitability research, habitat imagery, and existing inventory systems. Food preparation, recreation, and everyday comfort offered promising directions, but stowage kept coming up across activities.
Cargo transfer bags (CTBs) densely packed aboard the ISS. Image: NASA.
A maintenance work area documented by SQuARE, an archaeological study of everyday life aboard the ISS. Image: NASA/ISSAP.
NASA’s 2025 habitability review describes crew moving accumulated stowage to access equipment or clear areas for work. Keeping inventory current also takes communication and updates when items change locations. [3]
SQuARE offered a concrete example of how spatial arrangements change in practice: a designated maintenance area served largely as storage and arrangement space during the study period. Its observed use differed from its intended function. [4]
Stowage became our opportunity to improve how everyday space supports work. We grounded our inquiry in existing orbital practices while considering how those needs might carry into near-term commercial habitats.
Journey mapping
Finding a tool depends on searches, records, and other people
Keeping track has long been hard: over the life of the ISS program, about 13,500 tracked items have been reported lost [10]. NASA has documented extended searches, delayed activities, and crew frustration when tools and supplies cannot be found, and some items take days or weeks to find [13]. Stowage and inventory upkeep also take crew and ground-team time [5], and crew time is costly: NASA charges commercial users $130,000 an hour for it [14].
Crew member accessing cargo transfer bags. Image: NASA.
A "wanted" poster combines identification, location clues, and use context to support a search. Source: Fink and Berger, slide 9.
An illustrative journey reconstructed from secondary research, rather than a directly observed maintenance episode.
Finding the wrench drew on several channels at once: searching likely work areas, asking crew and ground, checking procedures and IMS records, and RFID or barcode lookup. When those failed, the work waited or found a workaround.
After use, the journey splits. The wrench can go back to its assigned stowage, or stay staged near the work or elsewhere until someone resolves it. The sources we found said little about that second branch, so I flagged it as a likely point of failure: a tool left there could become hard to find, or stay secured while its purpose and expected return went unclear.
We took both into expert interviews: how people find things, and what happens to items set aside during work.
Formative expert inquiry
Supporting life in orbit takes more than tracking objects
We interviewed one former astronaut and three specialists in commercial habitat design, aerospace research, and aerospace logistics. We asked how people find, use, and return items, alongside broader questions of comfort, privacy, and engineering feasibility. A space architect advised us as we began ideating and later reviewed the concept, examining sensing, legibility, and physical access.
Commercial habitats raise expectations for comfort, privacy, and control
Design executive in commercial space habitats · E1
The executive described commercial habitats moving from purely functional engineering toward more hospitable spaces, for occupants with months of training rather than whole careers. They also raised privacy and intellectual property: a commercial station may work like a multi-tenant office, where sharing a habitat does not mean sharing access to personal belongings or proprietary work.
On storage, they separated deep stowage from the equipment gathered for a procedure. That kit has to be treated as stored while the work is under way, then returned or handed to whoever picks the task up. They doubted that new occupants would reliably check items in and out by hand.
Life in orbit requires adaptation beyond training
Former astronaut · E2
The astronaut described how training could not fully convey orienting, moving, and settling into life in orbit. Familiar layouts, color cues, and routines became practical supports as they adapted. Finding and stowing things happened many times a day for every crew member, and new crew lost items often while they adjusted.
They also improvised with available space, using an emptied cargo module as a temporary laboratory by crew agreement. Crewmates understood why something had been left in place because they knew each other well and labels showed whose it was. They expected habitats shared with less familiar people, such as visitors, to need explicit rules for what can be moved.
Reconfiguration adds capability—and complexity
Researcher in computational systems for aerospace · E3
The researcher's work on folding structures showed how a single mechanism could enable substantial changes in form while using packed volume efficiently.
They emphasized mechanical simplicity, repairability, and reducing the expertise needed to operate a system: the more complex something is, the more likely it is to break and the harder it is to fix in orbit. This framed flexibility as a tradeoff: additional configurations needed to justify their mechanisms, failure points, and operating effort.
A research concept using a single routed thread to fold a flat sheet into a 3D structure.
Tracking depends on clear responsibility for upkeep
Aerospace logistics specialist · E4
Drawing on ground-based aerospace operations, the specialist described moving tools out of personal toolboxes to where they were used, and agreeing on who maintains inventory information.
They described a tracking system that failed because its users had not been involved in defining it and would not enter data. Maintaining a record needed a clear place in people's everyday responsibilities.
Accumulated use tests sensing, legibility, and access
Space architect and habitat design engineer · E5
The architect challenged us to consider use over time: equipment accumulates, spaces are rearranged, and maintenance changes the working environment. They questioned how sensing would perform amid growing clutter, and what happens when power or sensing fails, as when crews power down equipment and shelter during solar events.
They also raised legibility for international crews: colors carry different meanings across cultures, and emergency lighting washes some of them out, so cues may need shape or pattern as well as color. Physical access and understandable organization had to hold up through everyday change and interruptions to digital support.
The U.S. Destiny module shortly after installation in February 2001 and in November 2022. Equipment and stowage reshape a workspace over time. Images: NASA.
Research synthesis & design opportunity
How might crews keep track of items as work moves them around the station?
Across these conversations, flexibility mattered only if people could use and maintain it within everyday routines. Physical handling, record upkeep, and permission all had to stay understandable as arrangements changed. Two points in the wrench journey needed the most help: finding an item, and knowing what to do with one set aside during work.
The research surfaced four opportunity areas:
Visual clarity: color, material, and finish that make a dense habitat readable from any orientation.
Tracking and temporary storage: linking items set aside during work to inventory records.
Ownership and privacy: showing whether an item is personal, shared, restricted, or free to move.
Dense, adaptable stowage: compact, foldable, modular systems for limited volume.
These areas span an item's whole journey through orbital stowage, from arrival to departure.
An item's journey through orbital stowage. Inventory tracking with IMS and REALM spans the stages up to return.
Responsibility for records had to be agreed by the people doing the work (E4), and access norms could vary across occupants and organizations (E1). We carried forward a design hypothesis: support those agreements rather than prescribe them.
Design constraints & heuristics
Every addition had to earn its volume, attention, and power
Orbit set four constraints on any design:
Limited volume: every object competes for habitable space. A new device has to earn its footprint, ideally by replacing several dedicated fixtures.
Disorientation: without a fixed up or down, the design has to stay legible and usable from any orientation.
Microgravity: a released object drifts, so items need to be secured and released with fast, one-handed actions.
Material restrictions: many common materials and attachments require flight safety review. The design has to rely on space-compatible materials.
We translated the research into four heuristics, which we kept in mind while ideating and later used to narrow the ideas:
Heuristic
Research basis
How we’d evaluate it
Passive over active: tracking should be part of the work itself, not added paperwork
Experts described the effort of separate check-in and check-out steps (E1) and resistance to data entry without clear responsibility (E4)
Count the explicit inputs a task requires, compared with current procedures
Legible over complex: reduce visual noise; use color, material, and light to make status clear
Desk research found the habitat hard to read; crews rely on familiar visual cues (E2); cues must work across languages and lighting (E5)
Test whether people correctly read status across lighting conditions, color vision, and language backgrounds
Omni-directional: support approach from any angle and different body sizes
Crews learn to orient through familiar layouts and visual cues (E2); tools need to stay within reach of the work
Test access and retention across body orientations in microgravity analogs
Mechanically encoded: obvious physical affordances; rely on geometry over failure-prone electronics
Experts stressed mechanical simplicity and repairability (E3) and asked what happens when power or sensing fails (E5)
Remove power and check what remains usable
The heuristics also pulled against each other. Passive tracking relied on sensing that mechanical encoding would avoid, legibility to others could expose what crews kept private, and an adaptable surface added mechanism that simplicity argued against. We weighed these trade-offs as the concept converged.
Ideation & storyboarding
We narrowed to storage boards that keep items in view where they're used
We began with open-ended ideation to get the ideas research had left us with out of our heads. It was also a a good exercise in mapping out potential, general design interventions across the station and it's spatial, physical, software, and socio-technical systems.
Black hole catch-all: gathers loose items and releases one on request. From a cohort workshop, with Malik Zhang.
Vending cubbies and curved wall stowage: items dispensed on request from stowage built into the habitat.
Free-flying robots that fetch, return, and hold tools as a third hand.
Search and scanning: an inventory display, scanners at module entrances, and guidance to an item.
Mapped onto an item's journey, most of these ideas worked on retrieval: getting an item to the person who needs it, and sometimes back again. Ideas that collected loose items couldn't tell a stray tool from one set aside mid-task. Building stowage into the habitat's surfaces meant designing at the scale of the whole station, beyond what we could build or test. Guided search depended on station-wide infrastructure and overlapped with what IMS and REALM were building toward, and scanning at module entrances asked too much manual effort. Each idea answered "where is it?"
We storyboarded two directions. The first followed a newly arrived crew member retrieving a tool kit: an intent-first search, a route calculated around station traffic and the person's own body, handhold-by-handhold guidance, and a container lighting up on arrival. The second started with an overpacked cargo transfer bag, where emergency tools are buried among miscellaneous goods and drift free when it is opened. It proposed storage boards placed throughout the habitat that keep everyday hand-held items in view, with ambient light to help someone find an item they've looked up.
Retrieval and wayfinding storyboard by Nono Sun.
Storage board storyboard, generated by me with ChatGPT. Adapted from co-designed team version for this case study.
Both answered "where is it?" Wayfinding raised questions of station-wide architecture and ergonomics that reached well beyond a capstone. We pursued the boards: they worked at a scale we could prototype, and they would hold items usefully before any network existed.
Concept development
Knowing where an item is doesn't tell us what it's doing there
Defining the workspace
Before committing to a form, we listed the affordances the board needed and the signifiers that might communicate them. The list drew on human factors, ergonomics, and accessibility, and on the difference between operational and permanent storage. Several signifiers were still undecided.
Two conditions shaped every consideration. In microgravity a released item drifts, so holding had to be secure and the board had to stay where it was placed. Crews would brace with one hand, approach from any orientation, and sometimes need the board to move with them.
I next focused on the pattern language for retention: which objects the workspace had to hold, and what each one required. We scoped the work to small, hand-held maintenance items: hand tools, inspection probes, fasteners, connectors, small replacement parts, and task cards. They range from roughly 3–25 cm long and from 10 g to 1 kg. These items share three traits:
They are regularly removed, moved, or borrowed.
They temporarily belong to one activity.
Their position and grouping carry meaning, as when fasteners are grouped in the order they were removed.
What is it doing here?
As that work went on, my concern shifted from finding items to keeping them from being misplaced in the first place. A tool set down mid-repair raises questions that finding it doesn't answer: what is it doing here, and is it safe to move?
Our research had pointed here already. The post-use branch of the journey map was the part our sources said least about. The astronaut's crewmates knew why something had been left in place because they knew each other well, and expected less familiar occupants to need rules (E2). The design executive described a procedure's kit as stored while work was under way, then handed to whoever picked the task up (E1).
Where our focus moved: from keeping items in view to the arrangements that form while work is under way. Long-term stowage, retrieval, and inventory tracking stayed as context.
Two-panel pair: "Known location" · "Missing context".
A board at the work keeps the wrench's location known, but not the work it belongs to. Scenario frames generated with ChatGPT for our final capstone presentation.
Our later storyboards made the gap concrete. They took the repair beyond a single board: a borrowed tool, a part contributed by another crew member, two boards sharing one activity. In each, sensing could tell us that an item had moved, but not whether the move was expected. The same empty slot could mean an approved loan or a problem. Closing out the work raised the same question: a board can't be cleared until someone knows what belonged to the work and what still has to come back.
I framed this as material workspace legibility: whether a physical arrangement lets a particular person tell what it means for the work. Legibility depends on who is looking. An arrangement can be obvious to the person who made it, partly clear to a teammate, and opaque to a newcomer, and the person who made it can forget. Three distinctions followed:
Present doesn't mean available. A tool on a paused repair may still be needed.
Nearby doesn't mean part of the same work.
Absent doesn't mean lost. A wrench on loan is gone and accounted for.
Lent to Activity B, the wrench keeps a link back to the repair it came from.
Inventory systems such as IMS and REALM record what an item is, where it is, and its status. Whether a particular arrangement is paused, on loan, or finished gets worked out through records, procedures, and conversation. NET was designed to carry part of that on the workspace itself. In commercial habitats, that could make a setup understandable to someone who doesn't know its owner or history. I wrote this framing up as a position paper for SpaceCHI 2026, alongside a poster I presented there with the team's permission and credit.
Two design problems
The framing split the affordance list into two design problems, shown in the table's last column:
Reliable restraint: keeping items secure, visible, and reachable with one hand.
Networked activity: keeping track of what each item belongs to as work moves between boards and people.
Networked activity uses three terms. A board is one physical unit. A workspace is the set of boards one activity uses, which can be a single board. The activity is the work itself, and each board serves one activity at a time.
Accessibility and guidance applied to both. So did one requirement, which our experts had pressed us on (E3, E5): before anything else, the board had to work as passive restraint, without power, sensing, or a network.
Jimmy and Nono led physical prototyping of restraint, and I was its primary tester. I developed the activity model.
Physical design & prototyping
Make room for varied items while keeping them accessible
Reliable restraint was a necessary starting point: if the board could not hold items through a pause, nothing it showed about them would matter.
Crews already restrain tools with fabric rolls, elastic loops, foam cutouts, Velcro, and cords. Each slot fits one tool, which works for a fixed kit but not for whatever a task brings to the board. Velcro and cords take any item, but fastening each one by hand is slow.
Tool restraint aboard the ISS: a fabric roll with elastic loops beside a foam cutout board. Image: Google Street View, April 2017.
We looked for surfaces that could hold varied shapes without a dedicated slot for each, drawing on precedents like workshop tool walls and an off-the-shelf brush stand with a slit silicone surface. Early prototypes tried foam blocks, bristles, cardboard cells, and slit silicone, then a numbered series of retention mechanisms, from an iris aperture to compliant hexagonal blocks. A 3D-printed hexagonal frame explored a form that could collapse flat.
Restraint explorations: precedents, early prototypes, the numbered prototype series, and a collapsible hexagonal frame.
Shake and vehicle-motion tests checked whether items stayed held, and I tested one-handed use underwater. All three were Earth-based stand-ins for microgravity. I also researched feasibility and how prototype materials would translate to space-grade ones. Each prototype in the numbered series tested a different way to hold items securely and release them with one hand.
As part of the materials research, I mapped each prototype material to one used or specified in spaceflight. Every one had a counterpart except the adaptive textile, so NET's material risk sits in a single layer, and that is where development would start.
From prototypes to a concept
Testing pointed to one surface for every item, rather than zones with different adhesion strategies. We specified that surface by what it had to do: adapt to any item's shape, then grip and release on demand. FibeRobo, an emerging heat-actuated fiber from MIT Media Lab [8], inspired the idea and served as a stand-in. It changes shape but doesn't grip, so any material that can, whether from soft robotics or another smart material, could take its place.
FibeRobo contracting when heated. It demonstrates shape change, not gripping. Video: MIT News.
We made the board itself a hexagon, a form that came out of the prototypes: it tiles like a honeycomb so boards can join, and its six sides let crews approach and brace from any angle. Containment was a microgravity requirement from the start, and the shake tests reinforced it. The flat face keeps every item visible, like a bulletin board, while expanding the board into a contained volume gives it a second mode closer to a bag, for deeper items.
From the prototype explorations to one board: one surface for every item, a hexagon that tiles and can be braced from any side, and containment the shake tests reinforced.
Pivot: from physical to AI prototyping
Physical prototypes could loosely test restraint on Earth, but not NET in use in microgravity. I looked at simulation, from Ansys Student and Abaqus to NASA's Trick, but doing it well would have taken more time than the capstone allowed. So I proposed a more pragmatic route: AI prototyping, to show the concept in use. That produced the concept video at the top of this page and the renders below.
The developed concept: a board that holds, expands, and senses
NET combines a rear frame, a hexagonal frame with grip slots on every side, adaptive textile restraint, transparent containment, and light strips. Every material has a counterpart used or specified in spaceflight, mapped above, except the adaptive textile, which has none yet. Its physical design is built around the behaviors below.
Boards joined on a habitat wall, one expanded into a contained volume with clear sides and cover.
A single board, its textile conforming around a repair kit.
Surface dynamically accommodates different object shapes
The device expands on demand into a contained volume while keeping contents visible
The images above show the capstone version of NET, before the strips gained an e-paper layer and the manual state dial.
NET's physical behaviors, light states, and the shapes on its dial.
NET's physical layers, from the front frame to the rear frame that anchors it.
Restrain varied shapes. The adaptive textile is a speculation: it would conform to each item, so nothing needs its own slot. Whether a fiber like FibeRobo can grip, not just change shape, is an open question, and one I've taken to the FibeRobo researchers as a future application for their work. The lab's newer woven textiles, which puff up and shrink on demand, are a closer starting point than the single fiber [16].
Expand while keeping contents visible. A transparent front cover with a semi-permeable opening lets items in and out while containing them in microgravity. Transparent silicone side gussets enclose the volume as the board expands. Visibility lets people inspect the physical arrangement directly.
Stabilize with one hand and work with the other. Grip slots set into the frame's six outer sides keep the place to brace separate from the area holding tools.
One hand braces with its fingers in a grip slot in the board's side while the other places a tool.
Position the workspace near the activity. The anchoring concept lets a board move between habitat surfaces in any orientation, not only walls; its rear frame anchors it to the surface. A body-mounted configuration is an additional direction for development. Attachment strength, reach, release force with bare and gloved hands, clearance, and the effects of a deployed board all require evaluation.
Boards anchor to any habitat surface: walls, ceiling, floor, or an angled panel.
Sense through clutter. No single sensing method is reliable in a cluttered habitat, the concern E5 raised. The board carries its own in its layers: contact sensing, passive RFID, and vision. System design below covers how they combine with what the station has.
Interaction walkthrough
Track the work, not just the items
This is the networked activity problem. Two crew members, Ana and Kofi, illustrate how NET is meant to work: Ana pauses a repair, and Kofi needs its wrench.
The lifecycle and the original light states come from our capstone. The interaction model below is what I developed after it, from provisional loans and item states to lights that rest dark and the at-rest display.
Today, nothing ties Ana's tools to the unfinished repair: they sit in a work bag or Velcroed nearby, and IMS records where the wrench is stowed, not that it is in use. Kofi searches as the wrench journey showed, and on finding it has to interrupt Ana or take it and hope. If it then goes missing, the crew may search together: one ISS commander spent 30 minutes on a tool that later turned up in a docked Soyuz [23]. NET keeps the wrench tied to the repair through each stage below.
Original capstone concept video. The walkthrough that follows moves past restraint to the lifecycle of use: establishing, pausing, sharing, and reconciling one repair.
States and signals
Tools change hands constantly, so the interaction could easily get in the way of the work. I designed it around one principle: no interface unless something is wrong [22]. The board and its light carry ordinary work, while NET tracks each item's state underneath. An item removed by its holder stays in use while the activity is active. It becomes unresolved if the activity pauses or closes without it, if a loan's return condition passes, or after a time window the crew sets.
The green states need no one's attention. Only an unresolved item shows anyone an interface, and only to the people involved. At Reconcile, items still in use are returned or transferred, and crews can tighten or loosen any of the triggers below.
How an item moves between in use, used elsewhere, unresolved, and resolved, and what triggers each change.
The light strips summarize the workspace, but only when someone needs it. At rest the lights are off and an e-paper layer under each strip holds the state, so a station full of boards doesn't glow all day. The lights wake on approach, a lookup, or a change, and an unresolved board stays softly lit, since drawing attention is its job. Each strip can light in segments, so it can also point: when someone approaches, the segment nearest an item that is out marks it. Whether a segment can point precisely enough to single out one item is untested. Dark at rest is normal: a board that doesn't wake on approach has lost power, and its e-paper still shows the last state. Item details and the activity phase remain available in the status view. The capstone defined three light states, which stood in for the borrowing controls we discussed but didn't design. I've since moved Reconcile from white to blue, so white can mean free to use:
Establish — Associate the workspace with the work
Ana positions NET near the repair and gathers the required tools and parts. In the proposed status view on a crew tablet, they associate the workspace with the repair and its responsible person. Setup is the one routine step that uses the status view. Items detected on the surface can then be reviewed as part of that setup.
This initial association supplies meaning that item detection alone cannot establish. Where available, a task or procedure record could provide a starting point; the crew would still need a way to correct the association.
Use — Register movement during ordinary handling
Tools move between Ana's hands and the workspace. The board's sensing detects relevant changes, while approach or query brings up further information.
The model separates an observed movement from its interpretation: sensing reports what it observed, the system may infer a likely explanation, and a person confirms it. Removing an item during an active repair does not, by itself, establish a loan, a transfer, or a loss. An item removed by its holder stays in use while the activity is active. It becomes unresolved if the activity pauses or closes without it, if a loan's return condition passes, or after a time window the crew sets. If the system lacks enough information to account for a change, it must make that uncertainty visible.
The green states need no one's attention. Only an unresolved item shows anyone an interface, and only to the people involved. At Reconcile, items still in use are returned or transferred, and crews can tighten or loosen any of the triggers below.
How an item moves between in use, used elsewhere, unresolved, and resolved, and what triggers each change.
Pause — Keep unfinished work understandable
When the repair pauses, the tools remain associated with it. The workspace view identifies the activity as paused and shows the items still needed for its completion.
A paused workspace can remain accounted for. "Paused" describes the activity; "accounted for" describes whether relevant item movements are understood. Neither establishes that every item is available to borrow.
Share — Let items move without stopping the work
Tools move between Ana's hands and the workspace. The board's sensing detects relevant changes, while approach or query brings up further information.
The model separates an observed movement from its interpretation: sensing reports what it observed, the system may infer a likely explanation, and a person confirms it. Removing an item during an active repair does not, by itself, establish a loan, a transfer, or a loss. An item removed by its holder stays in use while the activity is active. It becomes unresolved if the activity pauses or closes without it, if a loan's return condition passes, or after a time window the crew sets. If the system lacks enough information to account for a change, it must make that uncertainty visible.
The green states need no one's attention. Only an unresolved item shows anyone an interface, and only to the people involved. At Reconcile, items still in use are returned or transferred, and crews can tighten or loosen any of the triggers below.
NET concept visualization: tools can move between people while remaining connected to unfinished work.
Evaluation & next steps
Test whether clearer status reduces the total effort of keeping track
The project produced physical explorations, a developed workspace concept, and an interaction model spanning interruption, sharing, and return. Different artifacts addressed different questions.
Evidence
What it contributes
What remains unestablished
Desk research and expert inquiry
Operational context, crew practices, and constraints for the design.
How consistently those findings apply across future habitats and occupants.
Physical mockups
Exploration of fit, deformation, footprint, insertion, removal, and handling.
Reliable restraint in microgravity, a validated item range, and flight-compatible materials.
Scenario sketches and concept video
A representation of how items and activity relationships change over time.
Comprehension, coordination effort, and recovery during actual multi-user use.
Proposed borrowing controls
A concrete interface flow to evaluate permission and clarification.
Whether its benefits justify the input and interruption it introduces.
The next study would compare two-person maintenance tasks using existing restraint and inventory practices with a NET prototype. Both conditions would include an interruption, an agreed loan, an unexplained removal, and return or transfer. A simulated sensing layer could first isolate the interaction model, with its role made explicit in the study.
I would examine whether participants correctly interpret availability and responsibility, how quickly they resume work, what remains unresolved at closure, and how many requests, corrections, and coordination exchanges the system requires. Brief workload ratings and interviews would help explain the observed effort. The baseline should be defined with domain experts before making claims about operational benefit.
Physical development would evaluate one-handed placement and retrieval, mounting stability, reach, visibility, and release without power. Technical work would address sensing under clutter, inventory synchronization, and the adaptive material’s power, thermal, and durability requirements.
The four criteria remain the basis for judging progress: administrative effort, legibility, access across orientations, and understandable physical handling. Search-time or workload improvements remain hypotheses until that comparison is performed.
SpaceCHI 2026 & reflection
Carry everyday coordination into habitat design
I presented our team’s concept through a lightning talk and poster at SpaceCHI 2026, under the title “Designing a Portable Stowage Board for Interrupted Work in Microgravity.” The conference program documents the presentation at NASA Ames. 9
Presenting NET alongside work on orbital, lunar, and Martian environments reinforced my interest in the everyday coordination that a habitat must support: finding equipment, understanding what is already in use, and handing work across people.
My question leaving the conference was how these concerns could receive explicit attention during habitat development. Their relevance beyond orbit would need to be studied in those settings, where physical conditions and operating practices differ.
The project also sharpened my understanding of interaction design. In NET, the interaction extends across a person’s grip, an object’s placement, a shared agreement, and an inventory record. Each part affects whether the arrangement remains understandable when work changes.
I want to carry that question into focused prototyping and evaluation: can the clarity NET proposes justify the effort and complexity it introduces?
References and research notes
External references support the background and technical precedents. Expert accounts support the inquiry findings. The opportunity, design criteria, and NET behavior are the team’s interpretation and proposal; the borrowing controls are identified above as a next-prototype refinement.
8. Forman, J., et al. 2023. FibeRobo: Fabricating 4D Fiber Interfaces by Continuous Drawing of Temperature Tunable Liquid Crystal Elastomers. Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology, Article 19, 1–17. DOI · Researcher-hosted publication.
Primary project research, 2026: E1, design executive in commercial space habitats; E2, former astronaut; E3, researcher in computational systems for aerospace; E4, space logistics manager; E5, space architect and habitat design engineer, later concept review. Findings are paraphrased from project notes and transcripts. Participant identities and private transcript links are omitted from the public case study.
Project artifacts: The wrench journey, physical mockups, scenario sketches, system documentation, concept video, and SpaceCHI poster are project outputs. Images of these artifacts should identify their purpose and stage of development. External photographs and research figures should link to their individual source records, in addition to the references above.