The Strange Problem with Making Things Safer by Making Them Denser
What do a battery fire and a rooftop parking garage have in common?
At first glance, almost nothing. One belongs to the future of transport, the other to an old concrete landmark. But together they expose a quietly defining problem of modern engineering: we keep trying to stack more capability into the same space, and then we are surprised when risk becomes harder to see, harder to contain, and harder to explain.
The electric vehicle is supposed to be cleaner, simpler, and in many ways safer than the combustion car it replaces. Yet the story of EV fires shows that safety does not disappear when danger becomes rarer. It changes shape. From 2010 to mid 2023, there were 488 reported light duty EV fires globally, 393 of them confirmed as lithium ion battery fires. In 2020 and 2021, incidents spiked, largely because of manufacturing defects that triggered recalls in specific models. The overall picture is not one of runaway catastrophe. It is something more interesting and more difficult: a manageable risk that still demands respect.
Now imagine the garage that sat atop the New Haven Coliseum, spanning 360 feet across the roof, supported by 358 foot trusses spaced 60 feet apart, with parking built into the structure itself. It was not merely adjacent to the arena. It was part of the arena. That is what makes the image useful. The garage did not stand safely outside the system. It was integrated above it, carrying not only cars, but the assumptions of the entire design.
That is the deeper connection between EV fires and rooftop parking: both force us to confront what happens when risk is concentrated inside infrastructure that is supposed to make us feel confident. The challenge is not just preventing failure. It is learning how to design trust when the hazard is embedded in the very thing we rely on.
Risk Has Not Gone Away. It Has Been Architected.
Most people think of safety as a property of the object itself. A car is safe or unsafe. A building is stable or unstable. A battery either behaves or it does not. But the more advanced our systems become, the less useful that view is. Safety is often not a thing a component possesses. It is a relationship between components, usage patterns, maintenance, geometry, and response time.
That is why the EV fire story matters so much. A battery defect is not just a battery defect. In a densely integrated product, one small flaw can travel across manufacturing batches, recall systems, public perception, insurance models, and emergency response protocols. The incident rate can remain comparatively low while the social impact becomes enormous, because modern systems are networked amplifiers.
This is where the garage on top of the Coliseum becomes more than an architectural curiosity. It is a physical metaphor for how we build trust now. The garage was not hidden in a lot far away. It was perched above the main structure, made possible by long trusses and careful load distribution. In other words, the building practiced a kind of confidence in engineering: we can put heavy things on top of other heavy things if the math holds.
That logic is everywhere in contemporary life. We place lithium ion batteries under floors, behind panels, inside phones, inside scooters, inside warehouse robots, inside homes, inside grid storage systems. The promise is efficient density. The cost is that failure, when it comes, is no longer isolated. It becomes harder to reach, harder to cool, harder to inspect, and harder to explain to the public.
The modern safety problem is not only that things can fail. It is that we increasingly design systems in which failure is most dangerous precisely because success was so efficient.
This is the paradox. We seek compactness, speed, and elegance, but every inch of efficiency can also remove slack. Slack is what absorbs error. Slack is what gives firefighters access, engineers margin, and operators time. When we remove it, we gain performance and lose forgiveness.
The Real Question Is Not “Are EVs Safe?”
The better question is: safe relative to what, and under what conditions?
That is because risk is not binary. It is comparative, contextual, and often misunderstood through dramatic anecdotes. A single fire can dominate public attention, especially when batteries are involved, because battery fires feel new, opaque, and technologically intimate. But the relevant measure is not fear. It is the combination of frequency, severity, detectability, and containment.
This is why the reported numbers matter. Hundreds of EV fires across more than a decade, amid millions of vehicles sold, do not describe a system in collapse. They describe a system in transition. The spike in 2020 and 2021, tied to known manufacturing defects in specific models, is equally important because it shows something many public debates miss: most technological risk is not abstract or mystical. It is often specific, traceable, and fixable.
That should change how we think about innovation. We often imagine progress as the elimination of danger. In reality, progress usually means relocation of danger into more legible, governable forms. Gasoline cars carry fire risk too, but EVs alter the physics, the detection problem, the suppression problem, and the repair problem. The fire may be less common in some contexts, but when it appears, the response toolkit is different.
The lesson is not to panic. The lesson is to stop using outdated mental models. A gasoline fire is a familiar enemy. A lithium ion fire is a different creature. It may require different sensors, different materials, different spacing, different charging practices, and different recall discipline. The same is true for the rooftop garage metaphor. You would never judge the safety of a high load structure by asking only whether the trusses are strong. You would ask how the load is distributed, what happens when one element fails, how evacuation works, and whether the system can tolerate localized damage without collapsing.
That is the common thread: mature safety is a systems question, not a parts question.
Density Creates Wonder, But Also Blind Spots
One reason these examples resonate is that they are both products of a modern instinct: the desire to densify.
In the car industry, densification means more energy packed into less space, more software controlling more functions, more convenience, more range, more performance. In architecture, densification means stacking uses, compressing footprints, and making land more productive. The New Haven Coliseum garage embodied that ambition directly. Why waste separate land for parking if the structure can support vehicles overhead?
The logic is efficient, but it comes with an invisible tradeoff. The more concentrated a system becomes, the more its failure modes become entangled. A rooftop parking garage is not merely a garage. It is a structural dependency that affects the arena below. An EV battery is not merely a fuel tank. It is a high density energy system embedded in software, supply chains, thermal control, and public policy.
This produces a subtle shift in how responsibility must be organized. In simpler systems, one person or one department could own a hazard. In dense systems, responsibility fragments. The designer blames the supplier, the supplier blames the material, the operator blames the user, the regulator blames the oversight gap, and the public is left with the intuitive sense that nobody is fully in charge.
That fragmentation matters because trust erodes faster than engineering can repair it. A low incidence event can still become a high salience event when the public sees a failure as emblematic of hidden complexity. This is why transparent research, clear recall systems, and practical mitigations are not optional extras. They are the social infrastructure of technical systems.
Think of it like this: if the garage is the visible structure, trust is the hidden support beam. It is easy to overlook until it is missing.
A Better Mental Model: Margin, Visibility, and Recovery
If the old question was “Can we make it work?”, the new question should be “Can we make it work with enough margin that we can still recover when it does not?”
That yields a useful framework for thinking about modern safety, whether in EVs or in buildings.
1. Margin
Margin is the extra room that makes failure survivable. In architecture, it is load reserve, fire separation, and access paths. In EVs, it is thermal management, battery chemistry selection, manufacturing tolerances, and conservative charging protocols. Margin is not inefficiency. It is the price of resilience.
2. Visibility
A system is safer when anomalies are easier to detect early. Battery defects identified in recalls are a visibility story as much as a safety story. The same is true for a building with clear inspection pathways and accessible structural elements. Hidden systems are not inherently bad, but hidden failure modes are.
3. Recovery
No serious system can guarantee zero failure. What matters is whether failure can be contained, isolated, and repaired. A rooftop garage that can be inspected and retrofitted is preferable to one whose design makes intervention nearly impossible. An EV ecosystem that can rapidly identify risky batches and update standards is better than one that waits for scattered incidents to accumulate into panic.
Resilience is not the absence of risk. It is the presence of spare capacity, clear diagnosis, and fast recovery.
This framework also helps explain why public debate so often becomes distorted. People ask whether a technology is good or bad. But the more useful question is whether the system surrounding it has margin, visibility, and recovery. A low risk technology without those qualities can still become dangerous in practice. A higher risk technology with strong controls can be responsibly managed.
That is not an excuse for complacency. It is a demand for precision.
What We Owe the Future: Design for Honesty, Not Just Performance
The deepest connection between these two images is moral as much as technical. Both remind us that the future should not be judged only by how impressive it looks when everything works. It should be judged by how honestly it acknowledges its own failure modes.
That means the next wave of EV growth cannot depend solely on better marketing or more range. It must depend on better chemistry, better quality control, better incident reporting, and better fire suppression strategies. The growth from 2023, with over 14 million light duty EVs sold worldwide, is a sign of success, but success creates its own obligations. Scale is not proof that risk is solved. Scale is the moment when risk becomes socially consequential.
The same principle applies to architecture and infrastructure. When a building puts parking on the roof, or any heavy use atop another, it is making a promise that the load paths, inspection regimes, and emergency response plans are adequate for a world that will not always behave as expected. Great structures are not just feats of strength. They are feats of anticipated imperfection.
We should want that in our technology too. Not perfection. Honesty.
Honest systems tell us where the danger lives, what it costs to control it, and what kinds of mistakes they can survive. Dishonest systems pretend that because the hazard is packaged neatly, it is therefore gone. That is the oldest mistake in modern design: confusing concealment with safety.
Key Takeaways
Do not confuse low frequency with low importance. A rare EV fire can still demand major attention if it reveals a systemic flaw or exposes weak response infrastructure.
Think in systems, not parts. Safety depends on design, manufacturing, maintenance, operation, and emergency response working together.
Treat slack as a feature, not waste. Extra margin in structures, batteries, and procedures is what makes recovery possible when something goes wrong.
Ask where failure would go, not just whether failure might happen. The most important safety question is containment: what happens next, and who can intervene?
Favor visible, inspectable complexity over hidden complexity. The more a system concentrates energy or load, the more it needs transparency, monitoring, and clear lines of responsibility.
The Future Belongs to Systems That Can Admit Their Limits
The rooftop garage and the EV battery are both expressions of ambition. One uses trusses to hold cars above an arena. The other uses chemistry to store enormous amounts of energy in a compact, mobile package. Both are feats of compression. Both depend on engineering discipline to make that compression survivable.
And both quietly teach the same lesson: the most advanced systems are not those that eliminate danger, but those that can absorb danger without losing our trust.
That is the standard worth keeping in mind as electrification spreads and infrastructure grows denser. The question is not whether we can keep making things smaller, faster, and more integrated. We can. The question is whether we will continue to build enough margin, visibility, and recovery into those systems that when something goes wrong, the future does not break with it.
