01 / LOOK BACKWhat do we inherit?
Ancient spaces open a question.
Created for
Earth Partner Prize 2026
Earth Partner Prize 2026
A RECIPROCAL SHELTER STUDY / DYLAN SCHOFIELD
Rock
protects
rock.
Exploring geopolymer homes that could protect people and reduce our impact on Earth.Ancient questions. Future materials.
Enter the study ↓DYLAN SCHOFIELDA HOME TO RETURN TO / CONCEPTUAL ARTWORK
THE HUMAN COST / 2000–2019 · GLOBALLY
What should
shelter protect?
Major disasters linked to natural hazards have a human and economic cost. Better shelter is one part of the response.
- Reported deaths
- 1.23 million
- Reported economic losses
- Approx. US$2.97 trillion
Source: United Nations Office for Disaster Risk Reduction (UNDRR) and Centre for Research on the Epidemiology of Disasters (CRED), 2020. The Human Cost of Disasters: An Overview of the Last 20 Years (2000–2019) ↗ EM-DAT records.
Worldwide totals across major disasters linked to natural hazards, including hazards beyond the four shown in the film. Economic losses include more than housing. These figures do not measure the performance of the imagined home.
WHY I BEGAN / DYLAN SCHOFIELD
The next home.
The next generation.
Creative Director &
Proof-of-World Designer
For thirteen years, I’ve worked across marketing, design and technology, within companies, at an agency and independently—and I still freelance today.
Five years ago, I began a new career in housing. Today, I work full-time as a mortgage agent and real estate agent in my hometown of North Bay, Ontario, Canada, continuing my family’s legacy in homeownership.
Rock Protects Rock grew from the first clear connection between my independent research and my work in housing. It led me to ask what our homes and cities could become—for people and Earth.
I give people, businesses and ideas an identity others can recognise and understand. My independent creative practice brings research and imagination together in worlds people can explore, question and discuss. I use “proof of world” to describe making an idea tangible enough to examine—its possibilities, its assumptions and what still needs testing.
In my papers and The Teaching Hut, I explore questions about ancient materials, places and the survival of practical knowledge. Those interpretations remain open to scrutiny. Rock Protects Rock carries the curiosity forward: investigate better shelter, measure its effects on Earth, and make useful knowledge easier to pass on.
The writing behind the question ↗01 / THE QUESTION WE INHERIT
What if shelter
gave something back?
The first “rock” is a geopolymer home. The second is Earth. The question is whether a mineral-based shelter can protect people while doing less harm to the planet.
01 / LOOK BACK
02 / LOOK CLOSERWhat can we test?
Evidence gives possibility substance.
03 / LOOK AHEADWhat do we leave?
A home belongs to a living Earth.
Sacsayhuamán-inspired masonry connects ancient craft with an imagined future material. The ancient casting idea remains a hypothesis to investigate. These are conceptual images, not historical reconstructions or tested specimens. Explore the masonry question ↗
ANCIENT QUESTIONS. FUTURE MATERIALS.
A question can
outlast a stone.
Sudama Cave in Barabar, India, has a rounded inner chamber and polished rock-cut granite surfaces. Its third-century BCE inscription connects the cave to the Ajivikas. The space invites us to look closely at the builders’ craft and the needs it served.
In separate studies of Cusco and Pumapunku, Trevor Hawke and Joseph Davidovits and colleagues propose that some stones were manufactured. Those interpretations remain disputed. Similar forms alone cannot establish shared material, method or scale.
See the places and evidence ↗02 / WATCH THE STUDY
What we build.
What they inherit.
4 min 50 sec · A cinematic material study
Sound on when you choose.
Read the narration +
More than four thousand five hundred years. Still standing.
Between two thousand and twenty nineteen, major disasters linked to natural hazards claimed more than one point two million lives and caused nearly three trillion U.S. dollars in economic losses, globally. What should shelter protect?
Joseph Davidovits proposed that some pyramid stones were cast. The idea remains disputed, but it opens a question worth carrying forward.
Could geopolymers help us build homes that protect people while placing less pressure on Earth?
For thirteen years, I’ve worked across marketing, design and technology, within companies, at an agency and independently—and I still freelance today.
Five years ago, I began a new career in housing. Today, I work full-time as a mortgage agent and real estate agent in my hometown of North Bay, Ontario, Canada, continuing my family’s legacy in homeownership.
Rock Protects Rock grew from the first clear connection between my independent research and my work in housing. It led me to ask what our homes and cities could become—for people and Earth.
Mineral ingredients, a different binder. A possibility, not a promise.
Hurricanes.
Earthquakes.
Wildfires.
Floods.
What if more homes remained? Wind, shaking, heat and water demand different answers.
We need to test the whole building: the shell, the connections, and the foundations.
An imagined home begins a question. It is not proof of survival.
Shelter also owes something to the planet that provides it, so could local mineral ingredients help reduce emissions?
Could longer service mean less demolition, less waste, and fewer resources consumed?
Measure the full life cycle, test durability, and publish what works and what does not.
Rock protects rock. The first rock is our shelter. The second is Earth.
Now imagine that care extending beyond one home. Different forms, shared spaces, a whole neighbourhood.
Plan for wind, shaking, fire and water with safer routes and infrastructure.
Combine disaster records and today's data to improve models and warnings where possible.
Help people leave danger before impact, without promising any home will always survive.
Research homes that more people could return to.
This is our moonshot: learn from what endures, test what comes next, protect life first, and build for its return.
Film sources and image credits +
The opening combines photorealistic AI-generated illustrations with a research question. The imagined homes that follow are artistic concepts, not tested survival results.
4,500+ years. The Egyptian Ministry of Tourism and Antiquities dates Khufu’s reign to approximately 2589–2566 BCE.
1.23 million reported deaths and approximately US$2.97 trillion in economic losses, 2000–2019 · Globally. Both figures come from the same report and period: The Human Cost of Disasters: An Overview of the Last 20 Years (2000–2019), published in 2020 by the United Nations Office for Disaster Risk Reduction (UNDRR) and the Centre for Research on the Epidemiology of Disasters (CRED), using EM-DAT.
The report covers major disasters linked to natural hazards worldwide, including hazards beyond the four shown in this film. Narration rounds the figures to more than 1.2 million lives and nearly US$3 trillion. This is not a count since 2005 or through 2026. Economic losses extend beyond housing; deaths were not all caused by building collapse. The invented storm scene is illustrative, not a depiction of the full dataset. No displacement total or count of destroyed homes is asserted.
The contrast is an invitation to investigate, not a controlled comparison of ancient monuments and modern homes. Joseph Davidovits’s casting hypothesis is attributed and remains disputed.
Original illustrations. The pyramid and fictional storm-damaged neighbourhood are photorealistic AI-generated images developed from the project’s earlier original 3D composition guides. Natural materials, flatter surroundings and gentle editorial movement give these illustrations a photographic appearance. Native animated typography remains. No borrowed documentary photographs appear in these replacement scenes. They are illustrative artwork, not archaeological reconstructions, records of Hurricane Michael or validated damage simulations.
The personal inquiry +
The first-person passage is Dylan Schofield’s approved statement, read by the same synthetic stock narrator used throughout. It describes his creative background, the move into housing five years ago, and this project’s connection between independent research and his work. The Teaching Hut title question appears as an excerpt from Dylan’s essay.
His papers and essay are not peer reviewed; FNIZ remains unvalidated. These writings explain the origin of the artwork. They do not establish an ancient construction method or the performance of an imagined home.
Read the writing and dated author’s note ↗EXPLORE A POSSIBILITY
What if a home
could stay standing?
One imagined street. Two ways to build.
Scroll to reveal the aftermath, or move the divider yourself.
Could a geopolymer home protect people through an extreme storm—and reduce the waste of rebuilding? This artwork pictures the outcome we want research to investigate.
IMAGINED CATEGORY 5 SCENARIOAI concept artwork · Not a performance test
Conventional houseGeopolymer home concept
Drag the line or use the arrow keys.
THE OUTCOME WE IMAGINE
A home still standing.
The dome is shown undamaged by artistic choice. That is a research goal, not a demonstrated result. Shape alone cannot predict survival; the material, openings, connections, foundation and site all matter.
THE EVIDENCE THAT STARTED IT
A reported case.
A new question.
Our Florida case study records a conventional concrete dome that survived Hurricane Irma in 2017. It does not establish geopolymer construction or the Category 5 outcome imagined here.
See the real case and its limits ↗Four environments. One responsibility.
Hurricanes · Earthquakes · Wildfires · Floods
A RECIPROCAL IDEA
The Earth becomes
our shelter.
What we build from it should help keep it livable.
Minerals can become the walls around a life. The responsibility runs both ways: investigate how those walls can resist harm, and how making them can leave less harm behind. Geopolymer research opens possibilities. Each possibility asks something of us.
02 / FOUR ENVIRONMENTS, TWO DIRECTIONS OF CARE
Shelter works
both ways.
Each environment asks how a home can protect life—and what building that home asks of the Earth. Each pairing is an artistic connection; the evidence differs across hazards.
01
Hurricanes
Wind / Resources
01 / SHELTER FOR USImagined sceneHURRICANES
When the wind rises.
The hurricane scene asks what a home could carry through a storm. A reported Florida survival case gives the question a real-world starting point; the mineral shelter shown here remains imagined.
Evidence informing the questionReported building precedent
02 / CARE FOR EARTHImagined possibilityMATERIALS
Give minerals another life.
Mineral by-products offer possibilities for some alternative binders. Could a shelter give suitable materials a further use while reducing demand for newly extracted resources?
Evidence informing the questionMaterial research + artistic ambition
Where the environmental promise stands +
Environmental value depends on sourcing, processing, safe use and the full lifecycle. The material composition and resource savings of this imagined shelter have not been measured.
02
Earthquakes
Movement / Inheritance
01 / SHELTER FOR USImagined sceneEARTHQUAKES
When the ground moves.
The earthquake scene places the desire for a lasting home against moving ground. It asks what resilience means when the ground beneath a shelter becomes part of the danger.
Evidence informing the questionHazard science; imagined response
What the evidence can tell us +
USGS describes ground motion and ground failure as sources of earthquake damage. That hazard science does not test this dome. Its shape and an ingredient’s compressive strength cannot establish seismic performance.
02 / CARE FOR EARTHImagined possibilityLONG LIFE
Repair more. Replace less.
A home that remains useful invites care instead of repeated replacement. Could repair become part of what we value in the materials and places we inherit?
Evidence informing the questionRepair ambition; benefit unmeasured
03
Wildfires
Heat / Emissions
01 / SHELTER FOR USImagined sceneWILDFIRES
When the forest burns.
The wildfire scene asks whether mineral materials could contribute to a home that is harder to ignite. Research on heated specimens and on whole-building exposure gives the question substance.
Evidence informing the questionMaterial experiments + wildfire research
02 / CARE FOR EARTHImagined possibilityCARBON
Lower the heat we add.
The second image turns from the heat we face to the emissions we add. Could alternative binders reduce the climate impact of providing shelter?
Evidence informing the questionLifecycle research; no project footprint
Where the environmental promise stands +
Lifecycle research supports reductions for some assessed mixes. Activator production, curing, transport and equivalent useful service affect the result. No carbon saving has been calculated for this artwork’s home.
04
Floods
Inland flooding / Water
01 / SHELTER FOR USImagined sceneFLOODS
When water reaches home.
The same ground-level dome faces rainfall flooding on ordinary inland land. Could a tested mineral shell limit lasting damage and help a home be repaired after the water recedes?
Evidence informing the questionFlood-damage guidance; imagined ground-level home
What the evidence can tell us +
Flood-damage-resistant materials can reduce water damage and ease cleanup. Remaining standing, keeping the interior dry and being safe to occupy are different outcomes. Water pressure, debris, uplift, foundation erosion, openings and utilities still matter. None of these outcomes has been tested for this imagined home.
02 / CARE FOR EARTHImagined possibilityWATER
Care for the water cycle.
The water tank and living landscape ask what a home could give back to the water cycle. Alongside freshwater care, could a repairable shelter reduce the waste left after a flood?
Evidence informing the questionWater-system research + land-care ambition
Where elevation enters the picture +
SAME FORM / DIFFERENT SITE CONDITIONS
A separate adaptation.
Flood depth, flow, duration, debris and ground conditions determine the question. Elevation may be relevant inland as well as beside water; distance from a shoreline is not enough to decide.
Steel supports and geopolymer-based piers are possible materials to investigate. Their suitability would require evidence of structural performance and durability, including corrosion or reinforcement protection. Neither material is specified or tested here.
The main scene explores a ground-level home’s exposure and recovery. This alternative asks when raising the living space could be part of the response. Neither image establishes a safe flood height or a buildable foundation.
All four chapters are imagined situations. The pictured home has not been engineered or tested, and no protection against a particular event, category or magnitude is claimed. The Florida case does not establish performance in the other three hazards. Earthquakes belong to the broader resilience inquiry here; they are not presented as a climate-driven hazard. These scenes are not shelter-in-place guidance.
03 / THE FORM OF A QUESTION
A rounded shelter.
An open question.
The dome gives the work a recognisable form: a mineral shell enclosing everyday life. It is an artistic proposition, with no prescribed construction system.
01 / MATERIAL
What could stone become?
The mineral surface connects ancient curiosity with modern binders. No recipe, shell thickness or structural capacity is assigned to the pictured material.
02 / SHELTER
What lets a home endure?
The repeated silhouette carries one human need through four environments. The ground-level flood scene asks about exposure and recovery; an elevated alternative considers a different response. Their proportions are expressive, not calculated.
03 / RECIPROCITY
What does a home return?
A garden, a greenhouse, water storage and solar panels make care tangible. They stand for possibilities to investigate; their presence in an image does not establish self-sufficiency.
THE POSTER SERIES
Better homes.
A healthier Earth.
That is the goal. These seven posters ask how geopolymer research could help us reach it. Open any poster to read or share the full image.
Ancient stone. New questions.
Ancient builders left enduring places. Could their work inspire new questions about geopolymer materials today?
Explore the past. Test the possibilities.
Open full poster ↗Image context +
Ancient geopolymer interpretations remain disputed. Imagined chamber; not a historical reconstruction.
Can geopolymer build better homes?
Geopolymers are mineral-based binders. Could homes made with them protect people while using fewer resources?
Test the whole home. Measure its impact.
Open full poster ↗Image context +
Sacsayhuamán-inspired concept artwork. No ancient recipe or building performance is established.
Safer homes. Less harm to Earth.
A home should protect the people inside it and reduce its impact on the world outside. Can geopolymer help?
Compare carbon, resources and repair.
Open full poster ↗Image context +
These are research goals. Benefits depend on the material, the complete building and its lifetime.
Four environments. Two responsibilities. +
Hurricanes
For people. Could a geopolymer home withstand extreme wind?
For Earth. Could fewer damaged homes mean less rebuilding waste?
What needs testing +
Test the complete building: shell, openings, connections and foundation. No hurricane rating is claimed.
Earthquakes
For people. Could a geopolymer home stay together when the ground shakes?
For Earth. Could easier repair extend its life and reduce waste?
What needs testing +
Strength alone is not earthquake safety. Movement, reinforcement, connections and ground conditions matter.
Wildfires
For people. Could a geopolymer home better resist heat and embers?
For Earth. Could its full life leave a smaller carbon footprint?
What needs testing +
Test the whole assembly, including openings and insulation. Heat resistance does not establish lower carbon.
Floods
For people. Could a geopolymer home suffer less flood damage?
For Earth. Could safer materials keep harmful substances out of water?
What needs testing +
A standing shell may still flood. Test water entry, erosion, contamination and safe recovery at each site.
04 / LIVING WITH THE LAND
A home is more
than its shell.
Food. Water. Energy.
Three possibilities for a closer relationship with place.
A place to grow.
The greenhouse represents a relationship with food. How much could it provide through local seasons, and what would it need in heat, water and labour?
What the sky offers.
The water tank asks how rainfall could meet some of a home’s demand. Collection, storage and treatment are separate systems; the shell’s curved surface does not create extra rain.
Living within a budget.
The solar panels ask how demand could meet seasonal supply. Their presence is a research invitation, with no claim that the imagined home can operate off-grid.
Give the promise
something to answer to.
Four questions
for the next investigation.
01 / CARBON
Does this mix reduce emissions for the same useful service?
Compare a locally sourced mix and a realistic conventional alternative using equal structural function, expected service life and lifecycle boundaries. Include activator manufacture, curing, transport and end of life.
Your choices stay on this page. Nothing is collected.
THE VIRTUAL MATERIALS LAB
Take your place
at the workbench.
A dome, a smaller pod, a green water tank. Explore the making of our geopolymer miniature from the maker’s point of view. Select the objects on the desk to move through six stages.
01 / PREPAREConceptual lab · AI-generated imagery
01 / PREPARE
A clear bench.
A careful beginning.
Start with the protective equipment on the left of the desk.
0 of 6 stages explored
Before handling real materials, a workshop needs a documented formulation, the product safety sheets, appropriate protective equipment and a suitable place to work.
Small pod diameter4.7 cm
Combined mould capacity161 mL
A geometry estimate, not a mixing ratio or an approved batch size. The picture is illustrative and does not change scale. Pod diameter = 7/12 of the dome; each hemisphere’s volume = π × diameter³ ÷ 12. Check the selected material’s pour limits.
Metakaolin
Heat-treated kaolin clay: a reactive mineral starting point.
Prepared silicate solution
The specified activator, paired with that powder in a documented system.
A virtual learning experience. Images and motion are generated illustrations, not documentation of an actual experiment. The model and its recipe have not been physically tested. Exact ingredients, ratios, safe handling, pour limits and curing conditions must come from the selected material system.
Read the complete casting notes +
01 / CHOOSE THE SHAPE
Two moulds.
One familiar home.
Use two open, hemispherical mould cavities: one large dome and one smaller pod. Cast with their flat bases facing up, then turn the cured pieces over for display.
Geometric capacity only, before losses or a sample tile. Not a mixing ratio or a recommended batch size. Confirm the material’s permitted pour volume and depth before choosing a size.
Notice: doubling every dimension needs eight times the material. Small trials make it easier to learn with less waste.
02 / CHOOSE A DOCUMENTED MATERIAL
Know what is
in the mix.
Our preferred direction is a modern metakaolin-based geopolymer: a reactive mineral powder paired with its specified, ready-prepared silicate liquid. Metakaolin is heat-treated kaolin clay.
- Ask a geopolymer supplier or materials laboratory for a documented formulation suitable for small castings.
- Obtain its exact ingredient grades, mass ratios, working time, permitted casting thickness, curing schedule and safety data sheets.
- Have the workshop supervisor confirm the material, mould compatibility and handling arrangements before mixing.
Why this route? The ingredients and conditions can be recorded and repeated. Sodium silicate can also appear in modern geopolymer recipes; it is not a clean dividing line between Davidovits’s work and Fóti’s experiments.
No particular beginner kit or model-specific formulation has been verified here. This lesson outlines the workflow; it cannot replace the selected system’s instructions.
03 / PREPARE THE WORKBENCH
Measure first.
Make a small trial.
- Prepare compatible moulds in a support tray, dedicated mixing tools, a suitable balance and a labelled sample-tile mould.
- Use the eye, skin and dust protection specified by the safety sheets, with the workshop’s ventilation and spill arrangements in place.
- Weigh the specified components and follow the documented mixing order. Start with a small trial tile to check setting and demoulding before committing to both domes.
Record: material and batch identifiers, measured masses, room temperature and the time mixing begins.
Alkaline silicate liquids can cause serious skin and eye injury. Use dedicated equipment and collect residues for the disposal route specified by the product and workshop; keep them out of drains.
04 / TRANSFER INTO THE MOULDS
Let the mould
give the material form.
- Once the trial has established a workable process, prepare the two clean, supported moulds with the specified release treatment, if required.
- Fill within the material’s working time. Use the supplier’s placement and air-release method, then level the exposed flat bases.
- Label a small companion sample from the same batch. Leave the castings undisturbed for the prescribed cure.
Notice: mould texture becomes surface texture. The lesson uses ordinary open moulds; it does not demonstrate underwater casting or promise bubble-free results.
Do not improvise extra water, lime, ash or pigment to change the flow. Each addition changes the formulation being evaluated.
05 / ALLOW THE MATERIAL TO CURE
Hardening takes
more than patience.
- Maintain the selected system’s specified temperature, covering and curing time. A mix that needs heat needs appropriate dedicated equipment.
- Demould at the documented release stage, supporting the whole piece rather than pulling on details.
- Continue any required curing before finishing. Photograph cracks, surface deposits or other changes alongside the companion sample.
What this teaches: demoulding time and full cure are different milestones. A hard-looking surface does not establish strength or long-term durability.
There is no universal “24-hour cure” for every geopolymer. This walkthrough is not a validated curing protocol.
06 / ASSEMBLE AND OBSERVE
Give the question
a tangible form.
- Place the cured dome and pod beside each other on a reusable base. Add a small decorative green cylinder on the pod’s right.
- Represent the door and windows with paper details or a compatible surface finish. Keep the castings solid for this introductory model.
- Label the model with the actual material, date and process. Photograph it with your batch notes and companion sample.
What have you learned? How a documented material can take a mould’s shape—and which observations you would investigate next.
Keep this distinction with the display: a successful miniature does not prove ancient casting, lower life-cycle emissions or resistance to hurricanes, earthquakes, wildfires and floods.
Explore colour next ↗Why this method? Research and practical limits +
Build on material research.
Davidovits’s published work includes modern metakaolin and silicate systems as well as his separate ancient-stone hypotheses. A 2019 technical paper by Ralph Davidovits, Christine Pelegris and Joseph Davidovits compares eleven commercial metakaolins and reports markedly different reaction behaviour. Its heated laboratory test is not a beginner casting recipe.
That variability is why this workshop calls for a complete documented formulation. Fóti’s reported water-glass experiments remain a research lead; the quoted “2% lime,” underwater and bubble-free claims are not independently established in this lesson.
Metakaolin testing paper ↗ · Davidovits’s chemistry reference ↗
Keep making and evidence connected.
MIT’s 2008 teaching project explored the pyramid hypothesis by making and comparing materials. It was an investigation, not historical proof. The 2006 Barsoum, Ganguly and Hug microscopy study was published in the Journal of the American Ceramic Society; it was subsequently discussed in a Nature news report.
MIT project ↗ · Original microscopy study ↗
For a concrete handling example, this sodium-silicate supplier classifies its solution as corrosive. Other formulations have their own requirements; read the sheet for the product actually used. Example safety data sheet ↗
THE MOONSHOT / BEYOND ONE HOME
Protect life first.
Build for its return.
My larger hope is for people to leave danger in time—and have a home, a community and a living landscape to return to.
MANY FORMS / ONE SHARED PLACE
Geopolymer is a material family.
It doesn’t have to be a dome.
Could we make room for different ways of living—and prepare the whole neighbourhood for hurricanes, earthquakes, wildfires and floods? This vision explores homes, shared spaces and essential services together, with people and the living landscape at its centre.
Domes, courtyards, gentle vaults.
Explore different shapes for different households, with daylight, privacy, accessible entrances and room to adapt.
Three front doors. A flowing enclosure.
Three modern homes sit side by side beneath an oval, wave-like form. Could shared structure and outdoor space support both belonging and independence?
Could a wall turn a wave?
The lakeside home’s curling wall imagines redirecting incoming waves. It needs testing: redirecting a wave would not stop a rising lake, and could shift risk to neighbours.
The wave wall is one question within a wider investigation: wind and debris, ground shaking, heat and embers, and moving or rising water. Each place needs its own evidence about these hazards and how they interact. Explore the design questions ↗
Longer records.
Better questions.
Investigate ancient accounts, historical disasters and geological traces alongside today’s sensors, satellites and climate data. Test whether new models can improve our understanding of where danger may grow.
A warning must
lead to action.
Research clearer warnings, accessible routes and support to leave before impact where lead time allows. Include people who need transport, care or help receiving an alert. Protecting life comes first.
A home depends
on its neighbourhood.
Plan around all four hazards: hurricane winds and debris, earthquake shaking, wildfire spread and floodwater. Study streets, open space, evacuation routes and essential services together with the buildings. The goal is less harm, less rebuilding waste and a safe return.
Four hazards. Questions for the whole neighbourhood. +
Hurricanes
How should wind, debris and storm surge inform siting, evacuation routes and the continuity of essential services?
Earthquakes
How can ground-risk mapping, safer buildings and resilient utility networks reduce damage and speed recovery?
Wildfires
How can planning reduce ember exposure and fire spread while supporting evacuation and healthy ecosystems?
Floods
Where should water be stored or conveyed, and where should development be avoided, so one area’s protection does not increase another’s risk?
A stronger material. A better warning.
A place designed to care for both people and Earth.
A research goal, with different limits for each hazard. Earthquake early warning starts after a quake begins; reliable date-and-time prediction is not available. An intact shell does not establish a safe home to return to. Read the evidence and open questions ↗
A PLACE TO LIVE / A PALETTE TO BELONG TO
Strong doesn’t
have to mean grey.
Geopolymer architecture could carry the colours of a place, too. Mineral pigments and carefully tested finishes open up possibilities for streets that feel expressive, familiar and alive.
Tap a grey home in the foreground to reveal its colour. The homes behind it are already coloured.
01 · Terracotta02 · Lapis blue03 · Ochre04 · Sage green
An imagined colour study. The colours are artistic choices, not tested recipes or certified finishes. The grey surfaces are a visual starting point; real unpigmented mixes vary in colour.
What is supported—and what still needs testing? +
Pigments have been tested.
A 2022 study of coloured slag-based geopolymer concrete tested green and yellow inorganic pigments. Results changed with pigment dose, including effects on strength and permeability. It supports testing colour as part of the mix, not assuming every colourant is interchangeable. Read the study ↗
LANXESS also reports collaborative trials of iron-oxide pigments in cement-free concrete. These are supplier-reported results for particular systems. Read the supplier report ↗
No catastrophe-proof film is established here.
KEIM Concretal-W is one example of a mineral facade coating. Its data sheet excludes weather-exposed horizontal and inclined surfaces. That is a useful reminder to check the whole application, not just the material name. Compatibility with the pictured geopolymer homes has not been established. Technical data sheet ↗
We have not verified a permanent film that makes a building safe in hurricanes, earthquakes, wildfires or floods. Colour retention, surface protection and whole-building safety are different questions.
ABOUT THE WORK
Hope with
its eyes open.
The first “rock” is a geopolymer home. The second is Earth. The question is whether a mineral-based shelter can protect people while doing less harm to the planet. Rock Protects Rock grew from Dylan Schofield’s housing work, creative practice and independent writing about ancient construction. Read why he began. A repeated mineral form connects human safety with environmental responsibility.
HOW IT WAS MADE
Concept and creative direction by Dylan Schofield. The project combines digital art, a cinematic film and a research notebook.
Credits and AI disclosure +
Generative tools and AI assistance were used for imagery, narration, writing, research support and implementation. The film uses one synthetic stock voice with original sound design. The pictured chamber, specimen and homes are imagined; no physical installation or material test was carried out.
The notebook distinguishes research findings, supplier statements, historical hypotheses and artistic proposals.
About the artwork ↗THE QUESTION CONTINUES
What else could
the stones teach us?
Across places and centuries, people shaped stone around the needs of life. Looking closer at what they left may help us ask better questions about what we build next.
Giza · EgyptSacsayhuamán · PeruPumapunku · BoliviaBarabar · India
Illustrated places that inspire inquiry. Ancient geopolymer interpretations remain disputed; Barabar is documented as rock-cut granite. These sketches are artistic impressions.
Follow the ancient questions ↗ONE LAST QUESTION / A MORE DISTANT HORIZON
Where are we
on the horizon?
We are here: 2026. The urgent work is already in front of us. Research better shelter for the hazards communities face today, while preserving what we learn for the people who come next.
Explore the markers below. This is a map of evidence and possible action, not a countdown to catastrophe. The intervals are not to scale.
OBSERVATION / NOT A FUTURE DEADLINE
1.44°C above the pre-industrial average.
WMO’s January 2026 assessment placed the 2025 global annual surface temperature at 1.44 ± 0.13°C above 1850–1900. This is an annual estimate, not the long-term Paris warming measure. It tells us about the climate we are living in; it does not date the next disaster.
WMO · 14 January 2026 ↗THE PRESENT / OUR STARTING POINT
Let local risk choose the first tests.
Start with communities exposed to hurricanes, earthquakes, wildfires or floods. Combine local hazard records and models with housing vulnerability, displacement risk and community priorities. Then ask which material and building tests could make the greatest useful difference.
The project is at the question-and-proposal stage. The pictured homes have not been built or validated. The next step is an independently tested, unoccupied prototype—not a promise of safety.
Explore the four testing priorities ↓2026–2030 / AN INVITATION, NOT AN AGREED SCHEDULE
From promising material to published evidence.
Bring communities, laboratories and building professionals together. Compare locally available mixes, test complete assemblies under defined hazards, measure their full environmental impacts, and publish failures as well as successes.
These are proposed research milestones, to be shaped with communities, independent laboratories and the relevant authorities.
The proposed research agenda ↗2030–2100+ / A RESPONSIBILITY ACROSS GENERATIONS
Leave the instructions with the inheritance.
A durable building is only part of the gift. Preserve readable methods, ingredients, units, test results, failures and repair knowledge. Keep multiple copies in different places and teach people how to use them.
This is a stewardship horizon, not a predicted collapse date or a claim that any archive lasts forever.
How could knowledge remain discoverable? ↓TENS OF THOUSANDS OF YEARS / MODEL ESTIMATE
An ice age is not a near-term countdown.
A 2016 modelling study estimated that even without human-caused climate change, the next glacial inception would be no earlier than about 50,000 years away. Added carbon could delay it further. This is a model-based estimate, not a date at which humanity should move underground.
PIK · study published in Nature, 2016 ↗NO SINGLE CLOCK
Extreme floods and solar storms do not belong at a fixed point on this line. Solar activity follows an approximately 11-year cycle, but that cycle does not date the next damaging storm. Earthquake risk can be assessed; the exact time, location and magnitude of a future large earthquake cannot currently be predicted. NASA / NOAA ↗ · USGS ↗
Could shelter ever need to go deeper? +
Could geopolymer learning contribute to earth-sheltered settlements—or even underground cities, if humanity ever needed to adapt? These remain separate research questions, not evidence for a universal refuge.
Extreme floods
Going underground can increase danger from water and trapped exits. Safer locations, higher ground, evacuation and groundwater behaviour come first. NWS flood guidance ↗
Long-term cold
Would a settlement have reliable energy, fresh air, food, daylight and healthy shared spaces? A mineral shell alone cannot provide a liveable city. This is a distant design question, not an imminent ice-age forecast.
Severe solar storms
Earth’s atmosphere and magnetic field already protect people on the ground from harmful flare radiation. Power, navigation and communication systems need resilience; burial alone would not ensure it. NASA ↗ · NOAA ↗
Any underground habitat would also depend on ground stability, seismic behaviour, water control, ventilation, radon management, safe exits and community wellbeing.
A PERSONAL NOTE / DYLAN SCHOFIELD
A discovery should
not have to be
lost twice.
My work around homes, my experience in design, and my curiosity about ancient construction keep bringing me back to one concern: if we do not innovate—or carefully investigate what may have been forgotten—those who come next may inherit the same vulnerabilities.
But discovering a useful material is not enough. If geopolymer research, or another approach, helps us build better, we also need to pass on how it works. The ingredients. The methods. The mistakes. The evidence. Knowledge should remain findable and understandable, even if the systems we use to store it change or disappear.
I do not have a loss-proof mechanism. I want us to ask the question now, while we can act on it.
Dylan Schofield
Artist’s perspective · a proposal for others to question and develop
Let the material
carry the method.
Could stone or ceramic records preserve diagrams, units, ingredients, safety limits and the evidence behind a recipe? A beautiful monument should also leave instructions someone can understand.
Many places.
Many ways to read.
Pair durable physical records and documented time capsules with open digital files, offline copies and archives in different places. Record how to find and decode them; one hidden capsule is a single point of failure.
Teach it.
Test it again.
Invite future custodians to copy, repair and independently reproduce the knowledge. Preservation needs people, languages and recurring care as well as a strong storage material.
Could an archive someday return from orbit? +
I also wonder about more speculative possibilities: an orbital archive designed to return knowledge periodically, perhaps every ten years. Could anything like that remain recoverable and safe? It would need a separate feasibility study covering reliable storage, orbital debris, controlled return and whether anyone could still find and read it. It remains a speculative question, with feasibility still to establish.
No archive is guaranteed loss-proof. The aim is to reduce loss through redundancy, readable records and renewal. These are artistic research proposals, not evidence that a particular ancient geopolymer technology was lost. Explore digital preservation practice · NDSA ↗
AN OPEN CALL / FROM POSSIBILITY TO RESPONSIBILITY
Let the data
choose where
we begin.
To residents, researchers, engineers, architects, builders, emergency planners, archivists and people who care: help take the question further. Start where local evidence and community needs show that better shelter could matter most.
Wind + impact
Start with: wind records, debris exposure, storm surge and vulnerable housing.
Investigate: the complete shell, openings, connections and foundation under defined wind and impact tests.
Shaking + ground
Start with: seismic hazard, soil conditions, existing buildings and evacuation needs.
Investigate: cyclic loading, ductility, connections and foundation behaviour—not just compressive strength.
Embers + heat
Start with: fire history, vegetation, wind, ember exposure and safe routes.
Investigate: whole assemblies, vents, glazing and joints alongside the material’s heat response.
Water + recovery
Start with: flood depth, flow, duration, drainage and displacement risk.
Investigate: water entry, debris impact, scour, contamination and recovery, with siting and elevation assessed separately.
THE FIRST USEFUL CONTRIBUTION
Bring a place.
Bring a question.
Publish what you learn.
- Define a local problem with the people who live there.
- Agree on measurable safety, cost and environmental comparisons with qualified professionals.
- Test unoccupied prototypes under the applicable approvals, independently review the results, and share the methods and limits openly.
Help turn this proposal into independent, openly shared research. A material is one part of safety: land use, infrastructure, warnings and evacuation remain essential. Material and hazard evidence ↗ · Approval routes ↗
Protect the people.
Protect the Earth.
Pass the knowledge on.