The Hidden Barrier to Wireless

When good glass goes bad, innovation finds a way through

Germany dreams of a digital future: smart cities, high-speed trains with seamless Wi-Fi, and 5G in every corner. But despite living in a world obsessed with innovation, the physical infrastructure around us (buildings, public transport, public spaces) remains out of sync with that vision. In other words, the network is evolving fast while the places we use have not fully caught up. 

5G-Advanced is rolling out, Release-19 is locking down features right now, and 6G is being shaped at the International Telecommunication Union (ITU) under the IMT-2030 banner. In other words, the next wave is arriving as a careful evolution, not a hard reset. That’s great news for innovators across the stack. At the same time, here is a complementary angle worth surfacing: it is not only about basebands, cores, or apps. Sometimes the building itself becomes part of the connectivity story. Efficient façades can quietly blunt mid-band RF indoors.

So alongside new features and architecture, it is helpful to ask a simple, practical question: does the signal make it through the envelope? 

Signal Lost: Why Buildings Still Break Connectivity 

The 5G to 6G era is an overlap, not a flip, but inside buildings the limiter is often physics, not features. The first place that physics bites is at the building envelope, where radio waves must cross from outside to inside, and in most cases the primary entry path is the window. 

However, most glass isn’t built for signals. It’s built for heat retention, great for heating bills, but terrible for wireless penetration. Especially low-E glazing, impose heavy outdoor-to-indoor (O2I) loss: recent wideband measurements report about 33.7 dB penetration loss at 6.75 GHz and 42.3 dB at 16.95 GHz, enough to erase much of indoor advantage unless the façade is addressed. Even regulators’ national coverage models understate the issue: Ofcom’s Connected Nations 2024 applies an average 10 dB building-entry loss for indoor predictions, while UK spectrum-sandbox measurements in modern double-glazed buildings observed 28–36 dB O2I loss depending on incidence angle.  

Technically, low-E glass is comprised of extremely thin layers of silver or other low-emissivity materials, great for thermal emissivity, but also highly reflective. Think of low-E glass like a thermos. A thermos keeps drinks hot or cold because a thin reflective lining bounces heat back toward the liquid, while the air gap between the shells slows heat flow. Low-E glazing works the same way. 

However, these windows exist for a reason: they cut heat transfer, so ripping them out can be counterproductive from a sustainability perspective. Whole-life-carbon guidance increasingly favors retrofit-first strategies that avoid the embodied carbon of premature replacement. For older, listed, or cash-strapped buildings, wholesale replacement isn’t realistic anyway, which is why practical retrofits to reduce entry loss are so important. 

nu glass: Turning Windows from RF Blockers into RF Gateways 

Switzerland’s nu glass, born out of research at EPFL, takes a refreshingly simple angle on indoor connectivity: don’t rip out high-efficiency glazing, tune it. Their portable laser process micro-etches existing low-E panes in trains so they become selectively transparent to mobile frequencies, done on site without removing the window, so thermal performance and the retrofit-first sustainability story stay intact. 

Technically, the laser writes an almost invisible pattern into the low-E coating, creating a controlled “frequency window” while the insulation remains effective, with each pane typically finished in about fifteen minutes using micro-lines around 25 µm wide, thin enough to be all but invisible.  

The idea grew from another EPFL project to boost train-glazing insulation and cut HVAC use, which is about a third of a train’s power. When a rail partner warned the new glass might block mobile signals, micro-lines in the coating became a way to let waves pass. 

The approach is passive and is credited with cutting electricity use by around 28 MWh per train per year (roughly the annual consumption of 20 people) while maintaining thermal comfort and avoiding extra HVAC load. 

By June 2025, nu glass had treated more than 20,000 windows. They also joined Deutsche Bahn’s Gigabit Train initiative, with summer demos in DB’s TrainLab alongside Ericsson, Telefónica, Vantage Towers, and Icomera. In France, starting in December 2024, SNCF Voyageurs and Île-de-France Mobilités also began laser micro-etching windows on Transilien Line N, where commuters had been frustrated by recurring mobile signal blackouts. 

If tuning the glass works on rails, imagine what it could unlock in offices, schools, hospitals, and factories, where heritage rules and tight budgets make replacements unrealistic, and where a quiet retrofit could turn energy-saving windows into RF gateways so existing networks finally reach the people, devices, and sensors that need them. 

Slices vs. Windows: Why Building Physics Sets Your Campus Network’s Performance 

Solutions like nu glass show what’s possible when we start looking at connectivity problems from a different angle. And while that approach has been proven on trains, the same thinking can be just as powerful in other settings, where materials-aware fixes tackle outdoor-to-indoor loss first, giving network features something real to build on. 

One of them? Campus networks. They promise private, high-performance wireless environments for industrial and commercial spaces. But here’s the reality: that performance only matters if the signal actually reaches the devices inside.  

This problem arises when companies choose the “Network Slice” model, where their private 5G network is logically separated but physically hosted on public mobile operator infrastructure. In this setup, the mobile provider supplies the radios, antennas, and core infrastructure, but the user has limited control over their placement or configuration. You can’t optimize the layout to compensate for signal-blocking materials. You can’t move the hardware closer to the pain points. Your signal is effectively boxed out by your building envelope.  

That’s why we need more ideas like this. Not just big infrastructure upgrades, but precise, retrofit-ready fixes that make the tech we already have work better in the environments we already use. Because making 5G and 6G work inside buildings shouldn’t always require tearing them apart. Looking ahead: as frequencies climb toward upper mid-band and sub-THz, untreated façades will be even more limiting, so engineered entry or dedicated indoor nodes should become a standard practice. 

You can talk latency, throughput, and edge computing, but none of it matters if the signal can’t get through the wall. In the end, it’s all about coverage. 

The Glass Barrier to Connectivity - And the Startups Breaking Through 

Sometimes, it’s not about building new networks. It’s about seeing the old world differently. The walls and windows around us aren’t just barriers; they’re opportunities to rethink how we bring innovation into the spaces we already live and work in. If we want the wireless future to really arrive, we need to stop looking only at what’s missing and start unlocking what’s already there. 

At xG-Incubator, we support founders rethinking connectivity from all angles, whether you’re fixing what’s broken or building what’s next.