In 2026 the Active Heat Exchanger (AHE) research programme has advanced considerably. We started as an engineering development project, found an engineering research question and then found a health research question. We address the problem of houses making their occupants sick, but we also solve difficult data problems highlighted by the COVID-19 pandemic.

Here is where we started:

I walked into the Edinburgh Hacklab one day in 2022 and I saw this arrangement of fans and tubes in the window:

This is Awaab Ishak, aged two years, and I saw this device could be a solution to the tragedy of Awaab's life.

Costa Talalaev was the designer, a physicist who runs Makerbee Ltd and we started to work together.

Two research questions

Having mastered the necessary mathematical topology design and high-precision 3D manufacturing skills, we were able to investigate:

1. Is it possible to control for three parameters (humidity, CO₂ and temperature) using only fan speed controls? We discovered we can, adding very little cost to turn our already-cheap retrofit ventilation device into a smart indoor climate management system; and
2. Can we provide data the scholarship identifies is lacking, in order to make policy decisions relating to indoor air quality and also disease transmission? So far it seems likely we can, by using AI to combine data across many devices. This has the potential to become a part of preventative health systems and also for active response during incidents of some kinds of disease and airborne pollution.

We have not completed this research by any means, but we are well on the way.

Awaab and housing

Awaab Ishak was two years old when he died of a mould-related illness in a damp flat in England. His death moved English politicians and from October 2025, Awaab’s Law requires English landlords to fix unhealthy homes.

The World Health Organisation estimates that 20–30 per cent of households in many EU countries have dampness problems. The UK is particularly exposed, with the oldest housing stock in Europe — probably the world, unsurprising given that 38 per cent of homes were built before 1946. 47 per cent of English homes have uninsulated walls, and of the roughly 8.5 million homes with solid walls across Britain, around 90 per cent remain uninsulated. Around 7 per cent of UK homes still lack double glazing, mostly the oldest and hardest to retrofit. In Scotland, 694,000 homes — 27 per cent of the stock — fall below the Tolerable Standard, and 1.2 million homes are below a C energy rating. The problem is particuarly bad in temperate oceanic climates.

In a flat like Awaab’s, there are three problems at once:

1. make the air warm with heating
2. keep the warm air (that you just paid to heat) inside with insulation
3. have good air ventilation to the cold outside, without throwing away the warm air

It is normal in the UK to solve just one or two of these problems in a retrofit by installing a new heating system or adding insulation, and while that reduces heating bills it also makes people sick, creating more Awaabs.

The initial solution

The Active Heat Exchanger was conceived when Costa Talalaev, a physicist who runs Makerbee Ltd in Edinburgh, had what seemed a simple problem: his old flat was damp. He couldn’t afford to tear it apart for ducting, he didn’t want to waste the heating energy he was paying for, and whatever he fitted had to work even with the draughts and leaks that are normal (often by design) in old Scottish buildings.

Using 3D printing and his lab measurement tools Costa created an origami-type design which worked for his flat. He then further proved it in the large communal room of the hacklab near his office, where he could monitor temperature and CO₂ over the web for months at a time. I saw the prototype running with its monitoring data on a screen, and joined the project on the spot.

Costa also founded Warm Edinburgh, a group of over 1,000 people working on tenement insulation and related problems, so he was hearing from tenants and landlords about humidity and black mould constantly. Existing ventilation solutions are designed for modern, well-sealed buildings with centralised ducting. They are inflexible, expensive, and they assume the building is airtight, which old buildings never are. Costa moved the solution away from mechanical rigidity into three components: cheap and highly controllable fans, intelligent control software, and comprehensive sensors. The result is cheap to produce, and even at hacklab scale the evidence from manufacturing PPE during the pandemic shows we can produce thousands of units per year before outsourcing to a factory for very high volumes.

Before showing where we are today, it helps to look at the principle of heat exchanging.

Traditional Passive Heat Exchange principles

The original Passive Heat Exchanger (PHE) built by Costa is a 30-centimetre box section tube filled with a honeycomb of hundreds of 5mm tubes. A diverter at one end splits the tubes so that half exit one way and half exit the other. You mount the whole thing through a wall, with a small fan at each end. Warm air leaving the building flows out through one set of tubes while cold air from outside flows in through the other set, and because the tube walls are extremely thin (as thin as a sheet of paper), heat transfers through them. The outgoing warm air heats the incoming cold air, so you get fresh air without losing your hot air.

This is the kind of PHE commonly found in German Passivhaus systems, and illustrates the general principle:

What Active Heat Exchange adds

In its simplest, passive form the device runs with constant fan speed and a manual control, a stripped down version of the initial prototype. The Active Heat Exchanger adds IoT sensors and controllable fans, and this is where something clever happens with the physics. Measuring airflow directly in a small tube is very difficult, but there is almost always a temperature difference between the two ends of the exchanger, and Costa worked out that by monitoring the temperature differential you can apply physics principles in equations to reliably infer the airflow rate and direction. That single measurement from a cheap temperature sensor, combined with humidity and CO₂ sensors, gives enough information to manage fan speed, compensate for wind pressure and maintain air quality automatically, all from a unit the size of a bathroom exhaust fan.

Each active unit optimises its own immediate environment using control theory, and by connecting multiples of these intelligent systems they can coordinate their activities for an entire building without needing a traditional centralised computer system.

Improving on the overseas normal

For the last thirty years the German Passivhaus standard has addressed all three of these problems at once, including a heat recovery ventilation system as the de facto standard. Heat recovery brings in fresh air from the outside, warming it with the stale air as it goes out, with ducting throughout the house. In contrast, the UK updated minimum heat recovery requirements in 2022, which do not apply to existing homes which need it the most. New build homes can be constructed as a sealed box where it is easy to do heat recovery, whereas old buildings are leaky, never designed for ducting and wind overwhelms ventilation fans and pushes all the heat out through leaks elsewhere.

The needs of any building are constantly changing. For example, Carbon Dioxide (CO₂) builds up when people are in the room and clears when they leave, humidity spikes when someone showers, and air pressure changes with the weather and wind. Managing these things is easier if the building is a sealed box, but there it is still a dynamic, changing situation ideally suited to Internet of Things device thinking.

We have interesting designs in progress (such as fitting into existing window bleed vents), but what we have now fits into a standard 100mm bathroom exhaust hole, exchanges 30 cubic metres of air per hour, and would cost perhaps £300 to an end consumer installing it in their house. It is remotely controlled and produces important data on what it is that people breathe in during most of their lives.

Where we know it works

There are some applications that are immediate and obvious, and do not need additional development because we have solved the engineering problem. The PHE is a drop-in replacement for an exhaust fan, fitting into the same 100mm hole. It works in old buildings because it doesn’t depend on the envelope being sealed. It handles wind because the control system adapts. And because the units are small and modular, they fit into spaces that centralised systems can’t reach.

Boats and live-in vans or caravans have serious condensation problems that most people don’t think about until they live aboard or travel in winter. A boat cabin or a campervan with people sleeping in it generates litres of moisture overnight, and without ventilation everything is soaked by morning. The second-generation PHE design fits into a panel mount for exactly this kind of installation, and in vehicles the passive version with constant low-speed fans is usually all you need.

The potential agricultural applications surprised us. In intensive horticulture, plants need CO₂ delivery, humidity control, and protection from temperature swings, and growers currently use gas heating for all of this at great expense. In animal husbandry, the needs are different (air drying, dust removal, disease vector control) but the AHE’s ability to manage airflow through narrow tubes suits both. The really interesting possibility is connecting plant and animal controlled environments via AHE, because their atmospheric needs are partly complementary: animals produce CO₂ and excess heat that plants can use, while plants produce oxygen. The narrow tubes also make it possible to divert gases through filters, including Zeolite systems that can capture methane. There are additional research questions here, but we have done enough investigation to believe this to be a promising field.

Where we are today

We are working, somewhat slowly, toward production-ready design and independent lab testing. Costa and I have been developing this since around 2022, and during the pandemic Makerbee demonstrated it could handle manufacturing at scale by producing 53,000 pieces of PPE on 3D printers in the lab. We’re confident the step from low-volume production to mass manufacture via extrusion factories is manageable once we have the design fully validated, but validation takes time and funding.

There is optimisation work ahead: improving efficiency per gram of weight, smoothing surfaces that meet airflow, adding automatic flaps against strong winds, detecting and controlling for insects and mould spores. We have been testing working units in real occupied spaces for over a year, and the core works well enough that we’re now focused on production engineering rather than proving the concept.

The goal is to make this cheap enough and simple enough to retrofit that it becomes the default way to ventilate a building, because that is what it will take to stop children dying of preventable mould-related illness in one of the wealthiest countries in the world.

The indoor data nobody has

The scientific scholarship acknowledges little is known about Indoor Air Quality (IAQ). The US Environmental Protection Agency has no monitoring network routinely measuring IAQ. The UK Parliamentary science office concluded in 2023 there are gaps in IAQ research around airflow and pollutant accumulation. A review of two decades of research found a lack of IAQ studies covering both residential and commercial environments. Those that do exist are mostly brief snapshots of single rooms, what the literature calls short-term monitoring bias.

Everything we do monitor is outdoors. Scotland’s entire air quality network measures ambient street-level pollution at fixed sites. The EU’s revised Ambient Air Quality Directive, in force since December 2024, strengthened outdoor monitoring but left indoor air outside its scope entirely. The EU’s SINPHONIE and OFFICAIR field campaigns measured indoor pollutants in schools and offices across member states, but these were time-limited research projects, not monitoring networks, and they did not cover homes.

The reason is cost, or at least that was the case until recent advances in reliable Internet of Things monitoring devices. Every indoor setting is unique in layout, occupancy, leaks, heating and more, so useful data requires sensors in thousands or millions of actual rooms. The parameters that matter for damp and ventilation — CO₂ as a tracer for ventilation efficiency, humidity, and airflow rate — are rarely measured together continuously in real occupied dwellings. These are the measurements needed to understand why buildings like Awaab’s flat become lethally damp, and what interventions actually work over time.

The active PHE changes this. Its control logic already requires continuous measurement of temperature differential, humidity, CO₂, and inferred airflow as the mechanism by which it operates. A network of deployed units would generate longitudinal indoor air quality data across a diverse sample of real homes: old tenements, new builds, boats, vans, across seasons and weather conditions. That kind of large-scale, long-term residential dataset is what the field most lacks, and it would arrive as a side effect of solving the ventilation problem rather than as an expensive research programme.

CO₂ and infection

Every breath you exhale contains CO₂, and so does every virus-laden aerosol an infected person releases into a room. Neither disperses instantly. CO₂ concentration is a practical proxy for the fraction of air in a room that has recently been inside someone else’s lungs — and therefore for how likely you are to inhale what someone ill exhaled a few minutes ago.

The use of CO₂ as a ventilation indicator dates to the 19th century, but it was formalised into infection risk modelling through the Wells-Riley model of airborne transmission, which shows that the likelihood of inhaling an infectious dose scales with the concentration of rebreathed air. During COVID this became practical public health guidance. SAGE advised that spaces with aerosol-generating activity should maintain CO₂ below 800ppm, and that CO₂ measurements could be used directly to infer airborne infection risk. Research in Nature Communications added something unexpected: elevated CO₂ at around 800ppm doesn’t merely indicate poor ventilation — it increases the aerostability of the virus itself, making SARS-CoV-2 more infectious in the air. CO₂ is not just a proxy for risk but a contributing factor.

The UK government distributed more than 386,000 CO₂ monitors to state-funded schools. Research monitoring 36 naturally ventilated classrooms found that airborne infection risks in winter were roughly double those in summer due to closed windows. The monitors worked: 96% of schools that used them could identify when to increase ventilation. Teachers learned, for the first time, what the air in their classrooms was actually doing.

Then as COVID receded and energy costs rose, most schools stopped monitoring. By late 2022 only 26% of classrooms were still tracking CO₂. The knowledge was demonstrated, the infrastructure was distributed, and then it was quietly abandoned because opening windows costs money in winter and nobody was telling people to be afraid any more.

The pandemic established that ventilation is an infection-control intervention, not a comfort issue, and that CO₂ is a practical tool for managing it in real time. That lesson produced no lasting infrastructure in homes, where most transmission actually happens. A network of active PHEs, continuously measuring CO₂ as part of their normal operation, would be that infrastructure — present regardless of whether a pandemic is in progress, acting on the data rather than displaying it.

So what next?

We need to develop the AI and data systems and deploy the Active Heat Exchanger to several hundred buildings, ideally a mix of homes and schools. This would then give us enough information to validate and refine the concept for city-scale rollouts. Some physical testing and certification is required, but we believe this to be relatively straightforward.

If you are interested, let us know. This is our passion project, but these vital topics of health and well-being should matter to local authorities in many parts of the world, and certainly in Scotland.