The Hidden Life of a Timber Building

Old timber buildings are often judged by what can be seen. A roof looks sound, there is no staining on the walls, the tatami is dry and the building smells as it should. Nothing appears to require immediate attention, and in many cases that may be an entirely accurate assessment.

It is also only a snapshot of one moment in the life of the building.

Traditional buildings are constantly interacting with their surroundings. Timber, earth, plaster and other permeable materials absorb and release moisture as temperature and humidity change. Rooms behave differently through the seasons, air moves through the structure, rain wets exterior surfaces, and drainage systems carry water away. Most of the time these processes are entirely normal, which is why one of the more difficult parts of caring for an old building is distinguishing ordinary environmental behaviour from the early stages of a developing problem.

For Kyoto, where tens of thousands of traditional wooden townhouses remain, that distinction has considerable practical importance. Preventative conservation depends not simply on knowing what a building looks like today, but on understanding how it normally behaves over time and recognising when that behaviour begins to change.

Moisture is not the enemy

Discussions about building monitoring can quickly become misleading when every high humidity or moisture reading is described as dangerous. Traditional buildings simply do not work that way.

Historic England’s conservation guidance explains that traditional materials such as timber, earth, brick and lime-based materials are often hygroscopic and permeable, meaning that they naturally absorb moisture from their surroundings and release it again when conditions allow.[1] This ability is part of the way traditional construction manages its internal environment, rather than a defect that needs to be eliminated.

The aim of conservation, therefore, is not to make an old timber building permanently “dry” according to an arbitrary modern standard. What matters is whether moisture becomes excessive, persistent or trapped in a place where the structure cannot dry normally.

When water cannot escape, the consequences can include fungal decay, mould, damage to finishes and the deterioration of timber. Historic England notes that structural timber kept dry can survive for centuries, while prolonged periods of elevated moisture create the conditions in which decay becomes possible.[2]

Japan’s Agency for Cultural Affairs makes a similar point in its guidance for cultural heritage buildings, warning that moist environments can encourage rot and harmful insects while recommending attention to matters such as underfloor ventilation and site drainage.[3]

The principle itself is straightforward. The difficult part is knowing what is happening inside the building before the consequences become obvious.

Why a single reading can mislead

Imagine taking a moisture or humidity reading beneath the floor of a machiya on a humid August afternoon. The number may be high, but that does not automatically mean there is a leak or a defect.

Perhaps that particular location behaves in much the same way every August. Outside humidity may have been high for several days, causing the materials beneath the floor to respond normally, and conditions may return to their usual range once the weather changes.

Now imagine that exactly the same reading appears suddenly during a dry period, in one isolated location, after months of otherwise stable measurements. The number is identical, but its meaning is very different.

This is why context is so important. Historic England warns that moisture assessment tools can be misunderstood and that both time and money may be wasted when damp is diagnosed incorrectly.[4] In an old building, the trend is often more useful than the isolated measurement.

What was normal last week? What happened in the same location last summer? How quickly did conditions change? What was happening outside at the time? Did the material return to its usual state afterwards?

A building with a history can answer questions that a single inspection cannot.

This is one of the fundamental ideas behind the KyoSensa platform. Rather than treating every reading as an isolated number, KyoSensa is designed to build an increasingly detailed record of how each monitored building behaves. Over time, that record can establish a baseline, or what KyoSensa describes as the building’s own fingerprint, against which future changes can be compared.

The practical benefit is important. An unusual reading does not have to be interpreted against a generic threshold alone. It can be considered in relation to the building’s own history, the behaviour of nearby sensors, recent weather and previous events. That creates the possibility of identifying meaningful changes earlier while reducing the false alarms that make poorly designed monitoring systems difficult to trust.

Kyoto’s weather is part of the building

The need for this kind of context becomes increasingly relevant when the weather itself is changing.

The Japan Meteorological Agency’s Climate Change in Japan 2025 report found a clear increase in the frequency of extreme heavy rainfall. Comparing the periods 1976–1985 and 2015–2024, the nationwide frequency of hourly rainfall of at least 50 mm increased by approximately 1.5 times, while hourly rainfall of at least 80 mm increased by approximately 1.7 times. Other measures of intense rainfall show similar increases, and JMA expects the national average frequency of extreme heavy rainfall to continue increasing under future warming scenarios.[5]

This does not mean that every machiya will suffer water damage. It does mean that the environmental conditions against which traditional buildings must be maintained are not static.

Roofing, gutters, external ground levels, drainage, service penetrations and subfloor ventilation all become particularly important when a building is exposed to periods of intense rain. After such an event, the most useful question is often not whether it rained, because everybody already knows that. The more important question is whether the building responded to that rain in the way it normally does.

From weather data to building behaviour

This is an area where continuous monitoring can provide considerably more useful information than a reading viewed on its own.

KyoSensa is designed to bring sensor data from inside the building together with local weather information, allowing changes in moisture and environmental conditions to be considered in the context of what was happening outside at the same time.

Suppose moisture rises in one monitored area during heavy rain. That is worth recording, but by itself it is not a diagnosis. If the reading returns to its established range shortly afterwards, it may simply represent the normal response of that part of the building to wet weather.

If moisture remains elevated long after rainfall has ended, however, the situation begins to look different. If the same location responds unusually during subsequent rain events while other parts of the building remain stable, a pattern is beginning to emerge that may justify investigation.

The eventual cause could be roof drainage, ground moisture, plumbing, exterior water ingress, poor ventilation or something else entirely. KyoSensa does not need to guess which one. Its purpose is to make the pattern visible so that the right person can investigate the cause with better information.

That distinction matters in heritage buildings, where an incorrect diagnosis can lead to unnecessary or inappropriate remedial work. Historic England warns that treating symptoms without identifying the source of damp can waste money and, in some circumstances, worsen the condition of a traditional building.[4]

More data does not replace diagnosis. Used properly, however, it can make diagnosis much better informed.

For owners, building managers and maintenance companies, there is another practical benefit. Instead of relying entirely on scheduled inspections to discover that something has changed, monitoring can help direct attention towards the places that are behaving unusually. This can make inspections more targeted and allow limited maintenance resources to be focused where they are most useful.

The parts of buildings people rarely see

The most valuable places to monitor are not necessarily the rooms that people occupy every day. Problems often begin in areas that receive very little routine attention: beneath floors, behind finishes, around roof structures, in service areas and other concealed parts of the building.

Traditional timber buildings are particularly sensitive to prolonged wetting in these locations. Historic England’s guidance on timber-framed buildings notes that problems can arise where moisture becomes trapped in wall cavities, around sill plates beneath floors, behind panelling and beneath impermeable finishes.[2]

These locations share an obvious difficulty: they are easy to ignore precisely because they are difficult to see. By the time deterioration becomes visible from the occupied part of the building, the underlying condition may have existed for some time.

Continuous monitoring creates the possibility of detecting an environmental change without repeatedly opening the building simply to look. In a heritage property, where unnecessary invasive investigation is itself something to avoid, that can be particularly valuable.

Discreet sensors can remain in locations that would otherwise receive only occasional attention and continue recording while the building is empty, overnight and during periods of severe weather. The resulting history can then be reviewed remotely through the KyoSensa platform rather than requiring somebody to be physically present simply to determine whether conditions have changed.

The technology does not remove the need for inspection. Instead, it can help answer a much more useful question: where should the inspection begin?

A baseline is more valuable than a universal threshold

There is another reason long-term records matter. Historic buildings vary enormously.

A threshold that makes sense in one location may be almost meaningless somewhere else. A shaded subfloor environment cannot reasonably be treated in exactly the same way as an upstairs office, while a heavily occupied commercial property will behave differently from a rarely used residence. Buildings that have stood for a century or more have also accumulated repairs, alterations, extensions and local quirks that are difficult to represent with a universal rule.

For this reason, KyoSensa’s longer-term purpose is not simply to become better at declaring readings “high” or “low”. It is to understand what normal looks like for the individual building.

As monitoring continues, the platform can build a picture of expected seasonal behaviour, responses to rainfall, drying and recovery patterns, and relationships between different monitored areas. Future readings can then be compared with that history to identify changes that are unusual for the building itself.

This is a deliberately conservative approach to heritage monitoring because it begins with the building rather than imposing an arbitrary model upon it.

It can also reduce one of the most obvious weaknesses of poorly designed alert systems: false alarms. If every humid summer afternoon generates an emergency warning, owners and managers will eventually learn to ignore the system. A useful monitoring platform must instead distinguish between conditions that are expected and changes that are unusual enough to deserve attention.

As that understanding improves, KyoSensa can provide an increasingly useful Building Fingerprint: a record of the patterns, responses and relationships that characterise the property over time. Sensors provide the measurements, but the longer-term value comes from understanding what those measurements mean in the context of that particular building.

Turning data into something people can actually use

Collecting data is relatively easy. Making it useful to the person responsible for a building is more difficult.

A building manager does not necessarily need to watch hundreds of sensor readings or interpret graphs every morning. What they need to know is whether the building is behaving normally, whether something has changed, where that change is occurring and whether it is significant enough to justify inspection.

KyoSensa is intended to provide that layer between the raw measurements and the person responsible for acting on them. Sensor readings can be considered alongside weather, seasonal behaviour, previous events and the history of the building, with significant deviations brought forward for review rather than expecting the customer to interpret every data point themselves.

For organisations responsible for more than one property, this becomes particularly useful. A maintenance company or property manager can monitor a group of buildings through the same platform, identify properties or areas requiring attention and retain a consistent historical record rather than relying on disconnected inspections, spreadsheets or individual recollections.

Regular reports also allow that history to become part of the maintenance record of the building. Over months and years, this can provide useful evidence of recurring conditions, the response to severe weather, the outcome of repairs and changes that might otherwise be forgotten between inspections.

The purpose is not to automate conservation judgement. It is to give the people making those judgements better evidence.

Earthquakes add another layer of context

Kyoto’s traditional houses also exist in a country where seismic risk cannot be separated from building management.

Research at Kyoto University’s Disaster Prevention Research Institute has included full-scale shaking tests of traditional Kyo-machiya in order to understand their seismic performance and evaluate reinforcement methods.[6]

Structural engineering and seismic reinforcement remain specialist disciplines, and environmental monitoring is not a substitute for either. There is, however, a useful role for contextual monitoring after significant local shaking.

An earthquake may justify inspection even when a building appears largely unaffected. If the environmental behaviour of the building also changes afterwards, perhaps through an unusual increase in moisture in a particular area, that information can help focus attention during the inspection.

KyoSensa is being designed so that local seismic information can be considered alongside the building’s monitoring history rather than existing as an unrelated piece of information in a separate system. After a significant event, the condition of the building before the shaking is already known and its behaviour afterwards can be compared with that established baseline.

The benefit is not dramatic, but it is practical. A post-earthquake inspection begins with more information than it would otherwise have had.

Monitoring repairs, not only detecting faults

Continuous monitoring can also remain useful after a problem has been found.

Suppose water ingress is identified and repaired. At what point is the affected part of the building actually back to normal?

Visible dryness is not always enough to answer that question. Historic England recommends monitoring during the drying of historic buildings and notes that the objective should be to return materials to an appropriate equilibrium rather than forcing them to dry as quickly as possible. Excessive heat or overly aggressive drying can itself damage historic timber and other materials.[7]

A continuous record can show whether conditions are gradually returning to the pattern that existed before the incident. This means the same system that helped identify an abnormal change can also provide evidence that a repair or intervention has had the intended effect.

For the owner or maintenance provider, this can be particularly useful. Instead of treating the completion of repair work as the end of the process, the building can be observed afterwards to see whether the underlying environmental behaviour actually recovers.

Monitoring therefore becomes more than an alarm system. It becomes part of the evidence used to assess whether an intervention has worked.

A building record that becomes more useful with age

The first month of monitoring can tell an owner something. The fifth year can tell considerably more.

Over time, a building accumulates a digital history of summers and winters, rainy periods and dry periods, unusual events, repairs, interventions and the way different parts of the structure responded to them. Patterns that are impossible to recognise during a short inspection may become much clearer when several years of data can be compared.

For an individual property, this history can improve maintenance planning and provide continuity when owners, contractors or building managers change. Instead of knowledge disappearing when one person leaves, part of the environmental history of the building remains available.

There is also a wider possibility if data from many buildings is managed responsibly and appropriately anonymised. Long-term records could eventually help researchers and conservation specialists investigate recurring patterns across traditional buildings.

Do particular types of underfloor construction show similar seasonal behaviour? How long do different areas take to recover after prolonged rainfall? Are particular interventions associated with better drying? Do some construction details repeatedly respond to extreme weather in similar ways? How does the response to an unusually severe rain event differ from ordinary seasonal variation?

These are not questions that can be answered from several weeks of monitoring. They require years of good-quality data before meaningful conclusions can be drawn.

That is precisely why collecting the data now matters.

The practical benefits of preventative monitoring

The value of a platform such as KyoSensa is not that it replaces the people who understand historic buildings. Its value lies in extending what those people are able to see.

Continuous monitoring can provide an early indication that conditions in a concealed part of a building are changing, while comparison with weather and historical behaviour can help determine whether the change is likely to be ordinary or unusual. A building-specific baseline can reduce unnecessary alerts, while remote access means a property does not have to be physically inspected simply to discover that nothing has changed.

When something does deserve attention, the history can help a technician or conservation specialist decide where to look first. After repairs, the same record can show whether conditions actually return towards their previous pattern. Across a portfolio of buildings, owners and maintenance organisations can gain a clearer view of which properties require attention instead of treating every building as equally urgent.

Over the longer term, the accumulation of this information creates something that periodic inspections alone cannot easily provide: continuity.

The building begins to acquire a memory.

That is arguably the most important benefit of the KyoSensa platform. Sensors are capable of measuring conditions at a particular moment, but the platform is intended to learn how those measurements relate to the behaviour of the building over months and years. As the historical record becomes richer, unusual changes can be interpreted against an increasingly detailed understanding of what came before.

Sensors measure. KyoSensa learns.

Technology that knows its place

Kyoto City’s cultural policy describes a city that has repeatedly renewed itself by combining tradition with new ideas. It also recognises pressures created by demographic change, shortages of people able to preserve cultural assets and rapid technological development.[8]

It would be easy to interpret this as an invitation to put technology everywhere, but that would miss the point.

The most appropriate technology for a historic building is often the technology that performs its job without changing the experience of the building itself. A machiya does not need to feel like a laboratory simply because it is being monitored.

KyoSensa does not need to make a traditional building feel modern. It needs to help the people responsible for that building understand it more clearly.

That means discreet sensors, careful interpretation, useful historical records and alerts that support human judgement rather than attempting to replace it. The technology should remain largely in the background while making the condition of the building easier to understand.

Watching what the eye cannot see

Kyoto is still losing Kyo-machiya at roughly 800 buildings a year.[9] Technology cannot solve the social, economic and development pressures behind all of those losses, and it would be unrealistic to suggest otherwise.

Physical deterioration, however, is one area where better information can help.

Traditional buildings have always communicated their condition through movement, smell, staining, sound, cracking and the judgement of people who know how to read them. Continuous monitoring adds another layer to that knowledge.

It can record changes while nobody is there. It can remember what conditions were six months or five years ago. It can connect a change inside the building with the rain that fell outside, identify a location that is behaving differently from the rest of the property, and show whether conditions returned to normal after a repair.

None of this is as dramatic as restoring a collapsed roof, replacing decayed structural timber or rescuing a building after serious water damage.

That is partly the point.

Preventative conservation is most successful when the dramatic intervention never becomes necessary. Its purpose is to understand what is changing while there is still time to investigate the cause and do something relatively ordinary about it.

For a building that has already survived a century or more, noticing those quiet changes may help it survive the next one.