Infrastructure Gaps 001
Small functions the world still leaves improvised.
A dog waiting outside a supermarket looks, at first, like a private decision.
The owner brought the dog. The owner chose to leave it outside. The owner went into the shop and intends to come back. Primary responsibility for that decision remains with the owner.
But there are two physically different situations that can look almost identical from the pavement.
In one, the dog is tied to whatever happens to be available: a railing, a bicycle rack, a lamp post, a piece of street furniture that was never designed for this purpose.
In the other, the site has intentionally provided something for the dog to use: a marked hook, a tether point, a small waiting bay, perhaps a place selected because it offers some shelter.
Survey evidence shows that some owners do leave dogs unattended outside shops, schools or other places, even though current site-level frequency and duration remain poorly measured.[1]
Those two situations may look similar from the pavement. They are not the same design situation.
The moment it becomes infrastructure
The difference is not whether the dog is happy, anxious or well behaved.
The difference is whether someone has intentionally created a function.
A bicycle rack can be used to secure a dog, but dog waiting is not its designed purpose. A retailer, landlord or site operator that installs a hook specifically for dogs has done something different: it has selected an object, a location and an intended use.
That is enough to make the arrangement designable.
Here, infrastructure does not mean a major public-works system. It means a deliberately provided physical function for repeated use. A bollard can be infrastructure. A bicycle rack can be infrastructure. A handrail can be infrastructure. Scale is not the important part. Function is.
Once a site intentionally offers a place for a dog to wait, questions that were previously accidental become deliberate design questions.
Why is the hook there rather than two metres away?
What can the leash reach?
Does the selected location put the dog in direct sun, in the path of bicycles, beside delivery traffic or directly against the flow of people entering the shop?
Can the fixture perform the job for which it is being offered?
None of those questions transfers ownership of the dog to the site. They simply recognise that an intentionally provided amenity can be evaluated on its own terms.
Design the waiting function before adding anything else
The physical questions come first.
Danish official guidance explicitly recognises a quick shop visit as one situation in which a dog may be tethered temporarily. The same guidance emphasises that the wait should be short, that the tether must allow the dog to stand, turn and lie down, and that weather, wind, shelter and shade may need to be considered.[2]
Those instructions are directed primarily at the person responsible for the dog. But they also reveal something important about the physical interface: movement, exposure and placement are not trivial details.
If a site deliberately provides the waiting location, those variables become legitimate things to inspect.
Start with movement.
A tether point is not just a metal loop. It creates a radius. Within that radius are walls, bicycles, doors, bollards, pedestrians, other dogs and possibly vehicles. The useful question is not simply whether the hook can hold a leash. It is whether the geometry of the whole waiting position makes ordinary movement possible without creating obvious hazards.
General tethering guidance from BC SPCA identifies entanglement around obstacles and proximity to high-traffic areas involving vehicles, people or other animals as foreseeable risks.[3] That guidance is not a supermarket design standard, and its numerical recommendations should not be imported as one. But the failure modes are relevant.
A waiting point placed beside a bicycle rack may expose the leash to wheels and frames. A point beside an automatic entrance may place the dog in continuous pedestrian flow. A point beside a loading area may be technically close to the shop while being a poor place to wait.
A designated point should also be judged as part of the entrance system rather than as an isolated fixture. The leash does not stay inside the footprint of the hook. It can cross a walking line, wrap around a post, or bring the dog into contact with someone who did not choose the interaction. That makes circulation part of the design problem. The aim is not to create a separate zone for every possible user conflict; it is to notice when the chosen waiting position turns ordinary movement through the site into a predictable source of interference.
There is also a difference between a fixture that merely exists and one that can stay fit for use. A bent hook, loose mounting, blocked waiting area or permanently cluttered approach can defeat a sensible original design. Even very simple infrastructure therefore has a maintenance question: who notices when it has stopped working as intended? The answer may be an occasional visual check rather than a formal inspection regime. The point is proportionality, not bureaucracy.
Then there is weather.
A location can be physically convenient and still be badly exposed. Shade can disappear as the sun moves. A wall can shelter from wind in one direction and amplify heat in another. The point is not that every waiting place requires a canopy, water station or engineered shelter. The point is that location itself has environmental consequences.
The surface matters too, but this is where the article should resist inventing standards. There is no need to declare one universal flooring material, one tether length or one spacing distance between dogs. The useful discipline is simpler: inspect whether the immediate environment defeats the intended function.
And then ask who has to maintain whatever is provided.
A hook may require almost nothing beyond an occasional check. A marked bay may need cleaning. A water bowl creates a refill and hygiene task. Shade structures can require maintenance. Every feature adds some operational burden.
That matters because the lowest-complexity answer may be the best one.
A simple, well-placed hook may be enough.
Or the correct answer may be no dedicated hook at all.
The first question is not what the waiting point can sense. It is whether the waiting point is well designed when no sensor exists at all.
Responsibility for the function, not custody of the dog
The owner remains primarily responsible for deciding whether this particular dog should wait outside, for how long, and under what conditions.
Providing a place to wait does not erase that.
But a useful design principle is:
Control makes responsibility possible; foreseeable reliance defines its scope.
If a site operator intentionally provides a dog hook, users can reasonably rely on the hook being fit for the function it is presented as performing.
That does not mean the provider becomes responsible for the dog's temperament, medical condition, training or the owner's decision to go shopping.
It means something narrower.
A dog hook does not make the shop responsible for the dog. It makes the provider responsible, in the design sense used here, for the hook as a dog hook.
The same logic applies to other site-provided features.
If a location is explicitly presented as shaded, the shade becomes part of what is being offered. If a feature is advertised as monitored, the monitoring becomes part of what users may reasonably rely on. The more active the system becomes, the more its own failure modes matter.
But responsibility does not rise automatically with the number of features.
A simple hook in a sensible location can create fewer problems than a more elaborate installation that is badly maintained.
This is a design-ethics distinction, not a claim that providing a dog-waiting amenity transfers custody or creates any particular legal liability.
And the relevant actor may not even be the retailer. The exterior space may be controlled by a landlord, shopping-centre operator or municipality. Responsibility for the provided feature follows the actor that actually provides, controls or represents it.
The correct answer may be to provide nothing
There is a strong objection to the entire exercise.
Maybe the dog should not be left outside in the first place.
That is a legitimate answer.
An owner can choose another arrangement. A site can allow dogs inside where that is lawful and operationally acceptable. A site can decide not to provide any dedicated waiting infrastructure at all.
Those possibilities should come before a campaign for better dog amenities.
There is another complication.
Formalising the waiting interface could change behaviour.
A clearly marked dog-waiting point might reduce improvised tethering in awkward locations. But it could also make owners more comfortable leaving dogs outside, perhaps more often or for longer. We do not currently know the direction or size of that effect.
That makes the intervention itself part of the research question.
Improving a waiting point is not automatically better if the improvement also normalises a behaviour that would otherwise happen less often.
This is why minimum intervention matters.
The framework does not require a site to climb from hook to shelter to sensors to some elaborate managed system.
It can stop immediately.
The correct answer may be:
- do not leave the dog outside;
- allow indoor access where feasible;
- provide no dedicated infrastructure;
- or install one simple fixture in the right place.
Minimum intervention includes the possibility of no intervention.
Only then ask whether information is missing
If physical design and ordinary observation solve the relevant problem, stop.
Only after that should a site ask whether there is missing information that could change a human decision — for example, elapsed waiting time or a local environmental condition that is not obvious from inside the shop.
That is the point at which telemetry becomes a legitimate question.
Not because sensors are available.
Because information remains missing.
Technology earns a place only when it reveals information that changes a human decision.
The infrastructure question
A dog hook does not make a shop responsible for a dog.
But once someone offers the hook as a dog hook, it is no longer just a piece of metal on a wall. It has an intended use.
That use can be supported well or badly. The fixture can be maintained, ignored, simplified or removed.
The owner still owns the decision to leave the dog.
The provider owns the design of what it chose to provide.
That is the moment the dog outside the supermarket becomes an infrastructure question.
Sources
[1] If Skadeförsäkring / YouGov. “Shoppingsugna hundägare oroliga för att hunden ska bli stulen.” 23 Nov 2016. Survey of 3,425 adults, including 823 dog owners. https://www.mynewsdesk.com/se/if_skadeforsakring/pressreleases/shoppingsugna-hundaegare-oroliga-foer-att-hunden-ska-bli-stulen-1663488
[2] Fødevarestyrelsen. “Hundesnoren skal bruges med omtanke – og kun midlertidigt.” 1 Jul 2025. https://foedevarestyrelsen.dk/nyheder/pressemeddelelser/2025/jul/hundesnoren-skal-bruges-med-omtanke-og-kun-midlertidigt
[3] BC SPCA. “Position Statement on Tethering of Dogs.” Approved Sep 2025. https://spca.bc.ca/programs-services/leaders-in-our-field/position-statements/position-statement-on-tethering-of-dogs/