HAAMDance research
Research through designWorking paper 01 / November 2026

Movement · attention · learning

Learning to move,
learning to notice.

Dance between bodies, places and screens.

Abstract

How does an explanation of movement become something a person can use while dancing? This ongoing investigation connects Kris Haamer’s dance-related practice, travel and creative work from 2004–2026 with a portfolio of learning tools developed at HAAM. A retrospective archive of 386 evidence records brings together classes, reflections, cultural encounters, films, references and software artifacts. Analysis identifies recurring questions about returning to practice, attending to breath and space, learning with other people, and translating movement through images and interfaces.

The project uses research through design to connect these questions with a lineage of prototypes: Dance Lab, Breaking and Mirror Coach, whose useful elements were consolidated into Keha in November 2026. The present contribution is an annotated research map and a programme of small, repeatable comparisons. The proposed practice studies have not yet been conducted; no learning-effectiveness result is claimed.

Keywords dance learning; embodied interaction; research through design; reflective practice; personal archives; movement interfaces

01

What helps movement become understanding?

A saved tutorial, a named step, a different camera angle and another person’s response offer different kinds of knowledge. The research follows what happens between encountering one of these explanations and returning to movement.

  1. RQ1 / Representation

    What becomes clearer when the view changes?

    Compare fixed video, mirrored views, slow loops, counts and earlier 3D studies. Look for changes in understanding of direction, timing, weight and support.

  2. RQ2 / Continuity

    What makes someone return to practice?

    Study how a small cue, a familiar phrase or a useful record helps practice continue between classes, places and devices.

  3. RQ3 / Attention & feedback

    Which cues can a dancer actually use?

    Follow breath, rhythm, spatial awareness and self-observation. Distinguish a person’s account, a manually recorded action and a measured signal.

  4. RQ4 / Context & connection

    What needs another person, a room or a cultural context?

    Ask how partners, teachers, community, music and the history of a movement change what an interface can explain.

02

Research through making, revisiting and moving

Research through design provides a way to develop knowledge by constructing and reflecting on artifacts [1]. Gaver’s account of annotated portfolios informs the decision to keep the prototypes together, recording the particular questions and judgments each makes available [2]. Movement-centred design also places the designer’s own bodily exploration within the inquiry [3]. Applying these approaches here is a research choice; the cited work does not validate these apps.

  1. 01ReconstructEncounter, role,
    source, uncertainty
  2. 02Make & annotateA prototype and
    the question it holds
  3. 03Return to practiceOne phrase,
    one comparison
  4. 04Reflect & reviseWhat changed?
    What remains open?
Figure 1. The proposed research loop. Archival reconstruction and prototype analysis are underway; the practice comparisons form the next phase.

Two connected bodies of material

The personal archive. The current map contains 70 journey records and 316 references. An entry can describe an experience, a plan, a document or another source for an existing event. Each records what happened, what it contributes and what is still uncertain. Duplicate submissions and related drafts are grouped where possible. Saved references and bookings retain their actual evidential scope.

The design archive. Repository history, implementation files, design notes and the November 2026 consolidation decision document how the software changed. Current GitHub source was reviewed on 16 November 2026. Earlier prototypes are read at preserved revisions, so their experimental features remain distinguishable from the current Keha implementation.

Analysis. For each episode, identify the role and action; retain any explicit reflection or attributed advice; connect it to a design question; then record a comparison that could complicate that interpretation. The author is both the subject of the retrospective account and the designer of the tools. This first-person position is useful for generating questions, and limits what can be inferred about other learners.

03

Four threads already visible

These are situated observations and interpretations from the archive. Their relationship to a prototype is a proposition to investigate.

Lisbon / 2012

Learning continues between classes

After a beginner Kizomba class, Kris wanted more frequent, intensive practice. The account raises a question about continuity: what helps someone return to a useful activity when a teacher is absent?

Read the class reflection ↗
Received guidance / 2015

Connection changes the explanation

Kizomba guidance emphasised checking tutorials with an instructor and practising connection with other people. This is advice received. Its presence in the archive invites a comparison between solo understanding and adjustments made in a partner or group setting.

Read the guidance summary ↗
Dancehall notes / 2024
“remember to breathe”

A list of movement vocabulary ends with a short attention cue. Another note names dancing and spatial awareness among things learned in Taiwan, while leaving their causal relationship open.

Read the class note ↗Read the spatial-awareness reflection ↗
Un-live Impromptu / 2020

The edit becomes part of the encounter

A completed collaboration joined contemporary dance, violin, audio-reactive visuals and film editing. Kris’s documented role includes editing and code-compatibility work. Team reflections concern mediation and presence.

Read the collaboration record ↗
04

A body of prototypes, one current home

The apps are artifacts within the inquiry. Their implementations show choices about explanation, feedback and practice. GitHub records a consolidation into Keha on 9 November 2026; the earlier applications retain their source history as references.

Current developmentA / Keha

A workbench beside ordinary practice.

Keha’s dance route brings together a reference vocabulary, an optional local camera mirror and loops from a video file chosen on the device. Its existing practice system supplies standing dance-coordination work and a shared history. The local workbench does not record, upload or score movement.

Deployment check · 16 November 2026: the public practice route is available. The newer dance workbench is implemented in the reviewed GitHub source and awaits publication in that deployment.

Question it holds. Does a short reference loop help someone return to a phrase and recall it after the screen is removed?

Imported vocabulary awaits instructor and source review. These are implemented practice affordances; their effects on learning remain to be studied.

Read the consolidation record ↗
Earlier studiesB / HAAM Dance Lab

Discovery → view → vocabulary → practice

Dance Lab assembled Pulse & Practice, Form Study, Encyclopedia and Mirror Coach into one workspace. Dance-up explored camera angles and anatomy; Tiaotiao organised names and cultural context.

Questions: representation and context. Which views help a learner understand a phrase? How can an interface preserve movement lineage?

The 3D studies and sample trend dashboards are preserved experiments. Their presence in the source does not mean they are active Keha features. Demonstration rankings were authored data.

Open the design record ↗Earlier public demo ↗
Earlier study / curriculum retainedC / HAAM Breaking

Counts, weight and contact with the floor

The Breaking prototype modelled movement through count-by-count support, weight and floor context. It added beat progression, tempo controls, manually recorded holds and an orbitable 3D stage. The movement records were retained in Keha.

Questions: representation and feedback. Can a view explain support without implying that the display has assessed a person’s technique?

Timers and “mark practised” actions are manual records. The 3D body is procedural illustration. Imported floorwork, freezes and power vocabulary requires appropriate instructor review and is not a self-teaching prescription.

Open the curriculum record ↗Earlier public demo ↗
Earlier native experimentD / Dance Mirror Coach

Making the source of feedback visible

The native scaffold explored camera capture, pose collection, beat grids and reference matching. Its documented synthetic fallback becomes an explicit boundary for the current research: a generated demonstration must remain distinguishable from an observed movement.

Question: attention and feedback. What would make a person able to tell where a piece of feedback came from?

Native pose algorithms remain pinned references. They are not claimed as ported or deployed Keha features. The synthetic fallback is excluded from the current Keha workbench.

Read the feedback-provenance finding ↗
GitHub provenance & preserved revisions

The repositories are private. These links are provided for authorised source review; public readers can use the evidence summaries above. Repository visibility has not been changed.

Dance Lab / preserved prototype
b740bb0 · GitHub source (private)
Breaking / preserved prototype
09142b5 · GitHub source (private)
Mirror Coach / preserved prototype
29411f9 · GitHub source (private)

Source status was checked against GitHub on 16 November 2026. Earlier public demos can differ from the preserved development snapshot.

05

Return the question to the body

The next phase uses small comparisons with a familiar, comfortable phrase. The unit of inquiry is a documented attempt and its reflection. The following protocols are proposed; no participant results are reported.

P1 / View & recall

Watch, move, put the screen away

Choose one familiar phrase. Compare a fixed reference with a short loop or mirrored view; vary their order across sessions. After each, describe the phrase without the display and revisit it later. Record remembered cues, uncertainty and any changes that can be described precisely.

Artifacts: Keha workbench; earlier Form Study as a comparison if appropriate. Evidence: version used, reference and section, immediate account, delayed recall, researcher notes.

P2 / A usable cue

One phrase, one remembered reminder

Revisit the breathing cue or one spatial cue with a familiar basic. Keep the phrase and music comparable. Alternate usual attention and the selected cue; describe differences in ease, continuity and attention, including no noticeable change. Repeat before drawing a broader conclusion.

Artifacts: an archive note and ordinary practice; a tool is optional. Evidence: exact cue, phrase, context and first-person observation.

P3 / People & place

Compare solo clarity with a shared encounter

At an ordinary class or social dance independently chosen by the researcher, note one question that seemed clear alone. Afterwards, describe an adjustment made with a willing partner or group. Keep another person’s advice attributed and distinguish it from the researcher’s interpretation.

Evidence: the adjustment, setting, roles, advice and remaining uncertainty. A written account is sufficient; recording others needs separate agreement.

Across the comparisons, keep an audit trail from the question to the source, the artifact version, the attempt and the interpretation. If additional participants are later involved, establish recruitment, consent, recording choices and a review process before the study begins.

Download the session protocol ↗
06

What this account can support

The current contribution is an archive reconstruction, a documented prototype lineage and a set of questions that can be examined in practice. The archive is selective: it reflects surviving records, partial searches and later recollection. Some dates, teachers, places and personal roles remain unresolved. A software feature, a saved reference and a learner’s account provide different kinds of evidence.

There is no completed participant evaluation of learning effectiveness in this project. Procedural animation, demonstration metrics and manual activity markers cannot establish measured technique or retained skill. The author’s combined role as learner, designer and analyst makes the account situated and potentially selective; future comparisons should retain negative, ambiguous and unchanged observations.

Public summaries provide a route into the evidence. Private correspondence, registration details, health logs, personal media and source documents remain outside the site. Reader reflections on the playful floor stay in that reader’s browser and are not a research submission.

The aim is a more precise account of when an interface helps someone notice something useful—and what happens when they start moving again.

References

  1. Zimmerman, J., Forlizzi, J., & Evenson, S. (2007). Research through design as a method for interaction design research in HCI. Proceedings of CHI 2007, 493–502.
  2. Gaver, W. (2012). What should we expect from research through design? Proceedings of CHI 2012, 937–946.
  3. Hummels, C., Overbeeke, K. C. J., & Klooster, S. (2007). Move to get moved: a search for methods, tools and knowledge to design for expressive and rich movement-based interaction. Personal and Ubiquitous Computing, 11, 677–690.

Project materials

Dance (research): public evidence archive; the linked public note summaries; and HAAM’s versioned design artifacts and consolidation records. All project interpretations are attributed to this ongoing investigation.

Cite this working page

Haamer, K. (2026). Learning to move, learning to notice: dance between bodies, places and screens (Version 0.2). HAAM. Working research page. dance-research.haam.co/research

This page is an ongoing design research account; it has not been peer reviewed.