Story Ideas for Robot Faces

These mini-graphic novel ideas connect the book's face-design projects to the real people who shaped facial-expression research, affective computing, social robotics, and human-robot interaction. They are written for grades 9–12, with enough conflict and discovery to become visual stories rather than illustrated biographies.

Dates and national descriptions identify historical context, not ownership of ideas: nearly every robot below was built by a multidisciplinary team. A full story should name collaborators, participating communities, and study limits.

Selection Criteria

Stories were selected for:

Relevance — a direct connection to facial anatomy, expression design, animation, affective computing, or human-robot interaction
Diversity — a range of cultures, identities, disciplines, and design philosophies
Inspiration — problems that invite young readers to observe, prototype, test, and revise
Drama — a clear question, obstacle, experiment, and consequence
Scientific honesty — room for uncertainty, ethical questions, competing theories, and the difference between displaying an emotion and inferring one

Story Ideas

1. The Smile in the Eyes — Duchenne Maps a Moving Face


Subject	Guillaume-Benjamin Duchenne (1806–1875), France
Theme	Discovery and the ethics of experimentation
Connection	Chapters 9–10: feature independence, eye and mouth movement, facial action coding
Panels	8 — a focused investigation with an essential ethical turn

In a nineteenth-century clinic, Duchenne combines electrical stimulation with the new technology of photography to isolate facial muscle actions, including the eye movement associated with a spontaneous-looking smile. The story then changes viewpoint: what did the subjects understand and consent to, and how should later scientists use knowledge produced by troubling methods?

Why this inspires: Students see that close observation can change a field, while scientific progress never excuses ignoring the people inside an experiment.

2. A Face Older Than Words — Darwin's Album of Emotion


Subject	Charles Darwin (1809–1882), England
Theme	Following evidence across species and cultures
Connection	Chapters 10–11: basic emotions, cross-cultural recognition, multimodal cues
Panels	8 — a linear trail from questions and photographs to a provocative book

Darwin gathers photographs, observations of children, and reports about people and animals while asking whether expression has an evolutionary history. Publication of The Expression of the Emotions in Man and Animals in 1872 opens a debate that still reaches every student who draws a surprised eyebrow or an angry mouth.

Why this inspires: A familiar face becomes evidence in a mystery that spans biology, culture, art, and technology.

3. The Valley on a Graph — Masahiro Mori's Warning


Subject	Masahiro Mori (1927–2025), Japan
Theme	The courage to question “more realistic is always better”
Connection	Chapter 11: anthropomorphism, uncanny valley, expressiveness versus complexity
Panels	8 — one counterintuitive idea, its long silence, and its later influence

In 1970, Mori sketches a curve suggesting that affinity can suddenly plunge when an almost-human robot moves or looks subtly wrong. The essay receives little attention at first, then becomes a design compass—and a hypothesis to test rather than a law to repeat—when androids and digital faces grow more lifelike.

Why this inspires: One carefully drawn question can outlive a laboratory and help students defend the deliberate simplicity of a pixel face.

4. Every Movement Has a Name — Paul Ekman Codes the Face


Subject	Paul Ekman (1934–2025), United States
Theme	Turning fleeting motion into a shared measurement system
Connection	Chapter 10: universal emotions, facial action coding, the core expression set
Panels	10 — discovery, field research, coding work, adoption, and scientific debate

Ekman and Wallace Friesen move from cross-cultural studies to the painstaking development of the Facial Action Coding System, describing visible movements as action units rather than relying only on emotion labels. The final panels show both its enormous influence and the continuing debate over how reliably a face alone reveals an inner emotional state.

Why this inspires: Students learn that a useful model can be powerful without being the final word.

5. Teaching a Camera to Find a Face — Takeo Kanade's Visual Puzzle


Subject	Takeo Kanade (born 1945), Japan and United States
Theme	Replacing intuition with testable computational models
Connection	Chapters 9 and 11: facial geometry, recognition accuracy, and the boundary between displaying and sensing expression
Panels	8 — a technical problem-to-demonstration arc

Kanade's computer-vision teams confront lighting, pose, resolution, and the enormous variety of human faces while building systems that can detect faces and analyze expression. The resolution comes with a crucial distinction for young roboticists: drawing readable eyes is not the same problem as inferring what a person feels.

Why this inspires: Geometry and statistics become tools that let a machine notice patterns its programmer could never list one by one.

6. The Face That Had to Develop — Minoru Asada and Affetto


Subject	Minoru Asada (born 1953), Japan
Theme	Building to understand how empathy develops
Connection	Chapters 9–11: facial mechanisms, affective development, uncanny-valley trade-offs
Panels	9 — a design mystery followed by multiple rebuilds and a reveal

Asada's team builds Affetto, a childlike research robot, then repeatedly redesigns its face and body to study expression, physical interaction, and the development of self-and-other understanding. Each revision raises the same question from a new angle: can a constructed system teach us how social understanding emerges without pretending the robot literally feels as a child does?

Why this inspires: Rebuilding is presented as the work of science, not as evidence that the first attempt failed.

7. No Master Map — Rodney Brooks Builds Behavior First


Subject	Rodney Brooks (born 1954), Australia and United States
Theme	Challenging the orthodoxy of top-down intelligence
Connection	Chapters 11–13: human-robot interaction, gaze, animation states, responsive behavior
Panels	8 — a clash of approaches culminating in Cog and the lab that supported Kismet

Brooks argues that intelligence can emerge from layers of situated behavior instead of a single internal master plan. At MIT, the humanoid Cog becomes a test bed for gaze, movement, imitation, and learning, while a new generation of researchers—including Cynthia Breazeal—pushes the idea into social and emotional interaction.

Why this inspires: Students discover that a small set of responsive rules can produce behavior richer than a giant script.

8. Emotion Belongs in the Machine — Rosalind Picard Opens a Field


Subject	Rosalind W. Picard (born 1962), United States
Theme	Arguing that intelligence without emotion is incomplete
Connection	Chapters 10–11: affective computing, valence-arousal, multimodal emotion cues
Panels	8 — a clear intellectual pivot from image computing to affective computing

Picard moves from signal processing and image research toward a controversial proposal: computers that interact with people must account for emotion. Her 1997 book helps name affective computing as a field, but the story ends with a design responsibility—systems should respond to affective signals carefully, not claim certainty about a person's private feelings.

Why this inspires: A researcher can change direction when an ignored part of the problem turns out to be the most human part.

9. The Face Is Not the Whole Story — Lisa Feldman Barrett Adds Context


Subject	Lisa Feldman Barrett (born 1963), Canada and United States
Theme	Correcting an appealing oversimplification with broader evidence
Connection	Chapters 10–11: expression ambiguity, cross-cultural recognition, multimodal cues
Panels	8 — claim, anomaly, evidence review, and a better design rule

Barrett and other scientists examine the popular claim that a small set of facial configurations can reveal emotion reliably in every context. As bodies, voices, scenes, goals, and cultures enter the frame, a more careful lesson emerges: robot designers may create readable signals, but they should not assume that the same shape lets a machine read a human mind.

Why this inspires: Students get permission to challenge a textbook model by asking what evidence it leaves outside the frame.

10. Meeting His Double — Hiroshi Ishiguro and the Geminoid


Subject	Hiroshi Ishiguro (born 1963), Japan
Theme	Using an android double to investigate presence and identity
Connection	Chapter 11: anthropomorphism, uncanny valley, gaze, timing, and human-robot interaction
Panels	10 — a full build-test-public-reaction arc with a philosophical turn

Ishiguro's laboratory constructs Geminoid HI, a teleoperated android made to resemble its creator, and uses it to probe what makes a remote body feel present. Tiny motions—blinks, breathing-like movement, gaze, and timing—become as important as sculpted skin, while observers' reactions reveal the limits of appearance alone.

Why this inspires: The strangest engineering project in the room can also be a serious experiment about what it means to be with another person.

11. A Gentler Face — Kerstin Dautenhahn and KASPAR


Subject	Kerstin Dautenhahn (born 1964), Germany, United Kingdom, and Canada
Theme	Designing with children and therapists, not merely for them
Connection	Chapters 10–11: minimal features, expression intensity, readability, inclusive HRI
Panels	9 — a classroom/therapy challenge, iterative redesign, and field deployment

Dautenhahn and the University of Hertfordshire team develop KASPAR, a child-sized humanoid whose deliberately simplified face and predictable behaviors support structured interaction with autistic children. Trials in schools show both promise and variation, keeping therapists, teachers, families, and children—not the robot—at the center of the outcome.

Why this inspires: Restraint, listening, and co-design can be more innovative than adding another flashy feature.

12. Help Without Touching — Maja Matarić's Socially Assistive Robots


Subject	Maja J. Matarić (born 1965), Serbia and United States
Theme	Helping people help themselves through social interaction
Connection	Chapters 11–14: HRI, feedback, personalization, state-based behavior, live interaction
Panels	10 — a career-spanning idea told through several real-world deployments

Matarić's teams ask how a robot can coach rehabilitation, learning, or healthy behavior through encouragement rather than physical force. From lab prototypes to homes, schools, and clinics, facial cues and body language must be personalized, evaluated over time, and designed to support—not replace—the human relationships around the user.

Why this inspires: Robotics becomes a way to expand people's agency rather than automate them out of the story.


Subject	Cynthia Breazeal (born 1967), United States
Theme	Building intelligence through responsive social interaction
Connection	Chapters 9–13: independent facial features, affective computing, gaze, animation states, timing, and HRI
Panels	12 — a full research arc from a daring question through construction, interaction, thesis, and legacy

Beginning in 1997 at the MIT Artificial Intelligence Laboratory, Breazeal and her collaborators build Kismet, an expressive robot head inspired by developmental psychology. Big eyes invite attention; movable ears, eyebrows, eyelids, lips, jaw, and head communicate changing internal states; gaze, vocalization, and carefully timed responses turn a pile of mechanisms and computers into a social feedback loop with a human caretaker.

The dramatic center is not “a robot becomes a baby,” but a harder engineering question: can a machine use the social scaffolding through which people help one another learn? The epilogue follows Kismet into the MIT Museum and traces how Breazeal's doctoral work helped establish social robotics and human-robot interaction as major fields.

Why this inspires: Kismet shows that lifelike connection can begin with observable design choices students can reproduce—where eyes look, how brows move, and when a response arrives.

14. Skin Deep? — David Hanson Chases the Expressive Android


Subject	David Hanson (born 1969), United States and Hong Kong
Theme	Crossing art, materials science, animation, and robotics
Connection	Chapters 9–12: facial anatomy, feature actuation, uncanny valley, expression interpolation
Panels	9 — a materials mystery, a public reveal, and a design-trade-off reckoning

Trained across art and robotics, Hanson develops flexible “Frubber” facial material and a family of highly humanlike robot heads. The story contrasts the mechanical ingenuity beneath a nuanced smile with the harder questions above the skin: Does realism improve communication? When does publicity outrun a robot's actual autonomy? What can a two-eye screen communicate more honestly?

Why this inspires: Art and engineering are shown as partners, while bold claims remain open to evidence and critique.

15. The Robot as Therapy Teammate — Ayanna Howard Designs for Access


Subject	Ayanna Howard (born 1972), United States
Theme	Turning advanced robotics toward children's agency and access
Connection	Chapters 10–14: expressive cues, engagement, feedback, interactive controls, inclusive design
Panels	9 — need, prototype, child-robot session, measurement, and redesign

After work on robots for extreme environments, Howard applies human-centered robotics to pediatric therapy and education. Interactive humanoids use expression, gesture, and encouragement as part of therapist-guided activities, while the research team measures engagement and adapts prompts instead of assuming that every child will respond the same way.

Why this inspires: A robotics career can travel from distant frontiers to a problem whose impact is visible in one child's growing confidence.

16. Two Eyes and a Bounce — Hideki Kozima's Keepon


Subject	Hideki Kozima (contemporary), Japan
Theme	Finding the minimum movement needed for shared attention
Connection	Chapters 10–12: minimal-feature research, gaze, rhythm, idle animation, multimodal cues
Panels	6 — one minimalist design and a compact observe-build-test insight

Kozima's team reduces a social robot to a small yellow body, two camera eyes, and four simple degrees of freedom. Keepon can look toward an object, share attention, rock, and bounce to rhythm; encounters with children reveal how little machinery may be needed to make attention and emotion legible without building a literal human face.

Why this inspires: A student with a tiny display can look at Keepon and say, “I have enough parts to try an important idea.”

17. The Machine That Watched Expressions — Rana el Kaliouby Builds Emotion AI


Subject	Rana el Kaliouby (born 1978), Egypt and United States
Theme	Making technology attend to nonverbal communication—and confronting its limits
Connection	Chapter 11: affective computing, facial action, multimodal cues, culture, and ethical HRI
Panels	10 — personal motivation, doctoral prototype, startup scale, and responsibility

El Kaliouby moves from research on an emotional-social intelligence aid toward systems that analyze facial and vocal signals at scale. Success creates the story's conflict: training data, culture, privacy, consent, and context all matter when software moves from recognizing a visible movement to assigning an emotion label.

Why this inspires: Students see both the excitement of taking research into the world and the duty to question what happens when it gets there.

18. The Mask the Machine Could See — Joy Buolamwini Audits the Coded Gaze


Subject	Joy Buolamwini (born 1989), Canada, Ghana, and United States
Theme	Exposing misleading performance claims through evidence
Connection	Chapters 9–11 and 16: face geometry, recognition accuracy, testing, design critique, and documentation
Panels	9 — a mystery, an audit, a public reveal, and a push for accountability

When facial-analysis software fails to detect Buolamwini's face but responds when she puts on a white mask, she follows the anomaly into the training data. Working with Timnit Gebru, she builds a more inclusive benchmark and reports large performance gaps across skin type and gender, turning one failed demo into a wider campaign for accountable AI.

Why this inspires: A bug that says “you are invisible” can become evidence strong enough to make an industry look again.

A Possible Reading Path

For a compact four-story sequence, start with:

The Valley on a Graph — why simplicity can be a strength
Kismet Learns the Social Dance — how facial features and timing create interaction
Two Eyes and a Bounce — how little a social signal may require
The Mask the Machine Could See — why every face technology must be tested inclusively

Together, the four stories move from design hypothesis to expressive system, minimal prototype, and ethical evaluation—the same cycle students use in the book's capstone project.

Research Anchors

These sources are useful starting points when expanding an idea into a fully referenced story:

How to Generate a Story

To turn any idea into a complete mini-graphic novel with a cover, narrative, panel prompts, and references, use:

/story-generator {Story Title} --panels {N}

The suggested count is a starting point; it can be overridden with --panels N. Image generation creates one cover plus N panels. At the story-generator skill's current estimate of about \(0.039 per image**, a 6-panel story is about **\)0.27, an 8-panel story about \(0.35**, and a 12-panel story about **\)0.51 when using its paid image-generation workflow.