Because it is not prudence which drives humanity forward or long term cohesive cooperation. What drives us, and in the process what forms the shapes of our cities, is necessity and convenience....
Because it is not prudence which drives humanity forward or long term cohesive cooperation. What drives us, and in the process what forms the shapes of our cities, is necessity and convenience.
Necessity and convenience begets necessities and conveniences, and not always in that order.
Forgive me if I misunderstand you, but I'm not sure this is a fair comment considering most cities in the United States were designed, in advance, as grids; or designed, in advance, as a series of...
Forgive me if I misunderstand you, but I'm not sure this is a fair comment considering most cities in the United States were designed, in advance, as grids; or designed, in advance, as a series of culs-de-sac. There is very much an element of prudence in all advanced design paradigms, even if the expectations designers had at the time of construction don't match up with those of residents a decade or century later. Plenty of streets, neighborhoods, and cities have been leveled and rebuilt in variously unnecessary, aesthetic ways; or theoretically productive, utilitarian ways; for the sake of what designers considered "the benefit of humanity." Humans are more than capable of working toward the "greater good," false or genuine, if they settle on a narrative that suits them.
Neither of the contemporary advanced city design patterns mentioned are necessarily more requisite nor convenient than a hexagonal street layout as far as municipal budgets are concerned, even if it seems more convenient to laypeople familiar with grids and culs-de-sac. As the video author points out, hexagons are an extremely efficient shape, theoretically enabling greater density and simpler, safer intersections. If nothing else, contemporary lifestyles do value minimizing costs, and that aligns with efficiency. The video doesn't go into a lot of depth, but the main reasons hexagons haven't been adopted, according to the presenter, are more to do with stylistic preferences of developers.
I don't have much of an opinion on hexagons, but from the extremely cursory glance I have into the idea from this video, it does seem like an idea worth more investigation. I'm aware of little modern research on such designs at scale; this fact doesn't preclude research or discussion from happening. The kind of hopelessness embedded in "well, humans are just lazy and stupid, so it can't work" comments doesn't strike me as useful.
I just looked at an image of a stacked hexagon and it looks cool but the streets would be limited by the length of each hexagon’s side. Also it would lead to much more intersections which does not...
I just looked at an image of a stacked hexagon and it looks cool but the streets would be limited by the length of each hexagon’s side. Also it would lead to much more intersections which does not look like a great idea either. I didn’t watch the video so excuse me if they discussed these points.
The video talks about three-way intersections, which are apparently a lot safer than four-way intersections. This makes sense to me intuitively (fewer theoretical collision points). I wonder if it...
The video talks about three-way intersections, which are apparently a lot safer than four-way intersections. This makes sense to me intuitively (fewer theoretical collision points). I wonder if it would be cheaper to operate traffic control devices on a system designed for such streets. Perhaps they could get away with being less complex if there are naturally fewer safety risks to the design. I'm not sure.
Do you feel it's important for streets to be long/straight? I wonder what traffic pattern flow would look like in a city based on hexagons but which has a variety of block sizes. Since hexagons are infinitely stackable, I don't see any reason why you can't just have a set of seven hexagons for a large public building. Some of the designs shown in the video seem to take this idea in stride.
Maybe I’m stupid but I don’t understand how hexagons are infinitely stackable while squares aren’t. To me this whole idea looks like a dream. Main cities are already built as is and nobody is...
Maybe I’m stupid but I don’t understand how hexagons are infinitely stackable while squares aren’t. To me this whole idea looks like a dream. Main cities are already built as is and nobody is going to rebuild them to have a hexagon layout. A lot of the cities not in America don’t even have this type of layout and are built chaotically.
As for streets being long and straight, that just seems more convenient to me to be able to orient yourself. For example if you turned to a wrong street while the layout of streets is perpendicular to each other, then you know that you can simply drive a bit further to the next intersection or two to correct your route, while in a hexagonal layout you would just get lost in all of the short streets with their unique names and their intersections.
You're not stupid. It's a concept we aren't used to. I had to draw out a bunch of hexagons to get a sense of why this might be beneficial and how its problems could be overcome. I'm skeptical on...
You're not stupid. It's a concept we aren't used to. I had to draw out a bunch of hexagons to get a sense of why this might be beneficial and how its problems could be overcome. I'm skeptical on some of the finer details, but it seems like something that could potentially be useful to society.
Squares are also infinitely stackable. Hexagons just have a better ratio of interior area to frontage; on a structural level, they can potentially maximize the amount of space used productively by a city and minimize the amount of space used unproductively. Bees build hexagonal honeycombs for this reason, not square or circular ones. Squares, like hexagons, tessellate to form a grid; but hexagons are closer to circles than squares are. Circles are "efficient" insofar as their perimeter is as small as possible relative to the area within that perimeter. By contrast, a square "wastes" space in that it requires more of a perimeter ("envelope") for a particular usable land area. In the context of a city, this means more resources must be allocated to unproductive connections (roads) between productive nodes (blocks). Equilateral triangles tessellate into hexagons and are less efficient in and of themselves. Octagons and most other polygons do not tessellate. Hexagons, as the highest n-sided polygon which tessellates, therefore resemble the "optimal" circle the most (reducing wasted space within a unit) while also forming a repeating pattern (reducing wasted space between units, enabling higher density).
I am tempted to speculate that a hexagonal layout seems inconvenient merely because it isn't common, not because it is necessarily inconvenient. As a repeated, ordered layout, it has a predictable set of rules if you know what the pattern is. A straight line is obviously very simple, but hexagons also offer parallelizable routes, they're just intersected by blocks. It's like if you're walking in a square grid with a series of minor streets which are interrupted by offset blocks, which is common in many square grids anyway; Philadelphia's grid, supposedly neat and orderly, has elements of these interruptions in many streets (follow Cypress Street or Locust Street west to east, for example). If one were to grow up in a hexagonal city, perhaps one's understanding of navigability would be fundamentally different, and a square grid would be mystifying at first. Their worldview would be literally different; the way they understand layout might rest on things that us 90-degree'ers never even consider. I think these things are relative and not absolute. (In addition, the increasing prevalence of technology in our day-to-day navigation may, in the more distant future, change the way we understand navigability as a concept. Case in point: the Hexagonal Efficient Coordinate System. As computers get better at using hexagons, a navigational system based on them may become optimal.)
Presumably a hexagonal city would have a few arterial corridors much like a square grid, depending on traffic needs. But from a city planning perspective, completely straight roads maximize speed at the expense of pedestrian safety (straight roads universally increase speeds, and higher speeds universally kill more pedestrians), and the complexity of four-way intersections may reduce traffic throughput more than necessary, depending on lane orientation, pedestrian crossings, etc. In theory, apparently, a hexagonal system could maybe alleviate some of these practical issues. By reducing straightaways, drivers of multi-ton killing machines would, on average, kill fewer unsuspecting pedestrians. (Many systems currently in use on straightaways to protect pedestrians increase wait times so much for pedestrians and automobiles alike that pedestrians often cross without a signal, increasing their risk of injury. The longer the wait time, the more likely this is to happen, so if we want to maximize safety for pedestrians we probably have to decrease average travel speeds for automobiles. Pedestrians crossing streets where drivers are less likely to get up to high speeds, including less straight or more interrupted streets, may be less likely to be injured by automobiles.)
A hexagonal grid's shorter road length per given area reduces distance traveled for the average trip, which could reduce a lot of travel time. And while I'm not traffic expert or mathematician, it's maybe possible that three-way intersections also have a higher safe (for automobile passengers) maximum throughput than four-way intersections (even though four-way intersections possibly/probably have a higher maximum throughput in general). Four incoming roads mean there are twelve conceivable directions traffic could flow (three per incoming road: left, right, or straight), a ratio of 1:3. In contrast, three incoming roads means there are only six conceivable directions traffic could flow, a ratio of 1:2. In each case, 50% of turns are quite dangerous (left turns). But three-way intersections might have fewer theoretical dangerous collision points if someone runs a red light, because at a 120-degree angle they would possibly be more likely to cause a rear-end or a side-swipe than a straight-on t-bone. If "right-on-red" is permitted in a four-way intersection, then 6/12 (50%) of possible routes through the intersection can safely occur at once. If "right-on-red" is permitted in a three-way intersection, then depending on how you define a "relatively less safe collision point" (like a 90-degree t-bone) vs. a "relatively more safe collision point" (like a rear-ending, or a side-swipe), the number of possible routes through the intersection that can semi-safely occur at once is 3/6 (50%, if only right turns are permitted, and never left turns); or 4/6 (67%, if right turns are permitted, and left turns are permitted from one incoming street at a time, the other two streets yielding according to a traffic signal). With a 120 degree angle between the incoming lanes of traffic, rather than 90 degrees, it's almost like merging on the highway, which we don't consider unsafe. If integrated into a small traffic circle (roundabout), it's possible that safety and throughput would be further improved, perhaps more so than in a comparable four-way intersection. I'm not convinced any of these designs are necessarily safer or more efficient in practice, but maybe they could be if implemented thoughtfully. I tried to find some conclusive studies on this, but the data itself seems occasionally contradictory in places that study conclusions favoring four-way intersections tend to hinge on; and most data on three-way intersections is about 90-degree t-intersections, which are a fundamentally different design; so I'm not sure what to think. Some search terms pull up radically different results, I guess because they aren't technically all studying exactly the same thing. This 2022 study from Pravinvongvuth & Matarage seems to be one of the more useful ones for this particular discussion. Interestingly, safety in 120-degree intersections seem to have been studied more for pedestrian flow, as in Chen et al. 2019, which also seems to favor 120-degree intersections in reducing traffic flow conflicts. But it's unclear to me how universal this is.
I imagine the implementation would be city-dependent. Some geographies might be harder to develop this pattern in than others, as is the case with square grids. Intuitively, travel times can be minimized by connecting nodes by a straight edge. However, a square grid introduces very inefficient diagonal travel routes. If a city's geography is highly linear, it's important for its travel times along that linear axis to be minimized, which is perhaps most easily done with a square grid. However, this makes diagonal travel times comparatively slow; an inefficiency. Manhattan might be an example of a somewhat linear district. But in the bigger picture, much travel in our cities is not linear along the two axes of a square (or does not have to be). Manhattan is not just a line but rather a rectangle; and residents have travel needs to places that are not purely north-south or east-west, like going to JFK Airport if you live in Washington Heights. Even within Manhattan itself, orthogonally diagonal travel can be slow relative to linear travel. Speculating: given a plane containing blocks of equally distributed population density traveling in random directions, one could maybe optimize average travel times by using a hexagonal grid instead of a square grid, because a set of n random node pairs on such a plane are more likely to include diagonal travel than purely linear travel. Thus a suitably large sample size of nodes would be more diagonal than not, and so would benefit from a layout which has relatively efficient diagonal travel. But I am not a mathematician. Perhaps there is something more complicated with travel times that I am not considering.
Again, lots of speculation from a non-professional going on here.
already built as is
That wasn't the suggestion of the video (have you watched it?) or anything I said. This is just an interesting design concept that we could use when building new cities or neighborhoods, or potentially for destroyed neighborhoods that are being redeveloped after a fire, flood, etc. Just because the status quo is what it is doesn't mean it is necessarily the best use of our extremely finite resources and money! The more inefficient our cities are, the more taxes we pay to maintain them (unnecessarily). It is always useful to think about alternatives.
I do appreciate the comments, because I'm not wedded to this concept by any means and am interested in other perspectives, but urban planning is the kind of field most susceptible to inertia and which benefits the most from ideas challenging the status quo. The hexagonal grids envisioned in the video do look a little dreamlike, but I don't see a great reason why they're automatically something to throw in the wastebin. Our cities today look absolutely nothing like the cities of 1000 or even 200 years ago; perhaps our cities 200 years from now will take after the humble bumblebee. Who's to say?
Because it is not prudence which drives humanity forward or long term cohesive cooperation. What drives us, and in the process what forms the shapes of our cities, is necessity and convenience.
Necessity and convenience begets necessities and conveniences, and not always in that order.
Forgive me if I misunderstand you, but I'm not sure this is a fair comment considering most cities in the United States were designed, in advance, as grids; or designed, in advance, as a series of culs-de-sac. There is very much an element of prudence in all advanced design paradigms, even if the expectations designers had at the time of construction don't match up with those of residents a decade or century later. Plenty of streets, neighborhoods, and cities have been leveled and rebuilt in variously unnecessary, aesthetic ways; or theoretically productive, utilitarian ways; for the sake of what designers considered "the benefit of humanity." Humans are more than capable of working toward the "greater good," false or genuine, if they settle on a narrative that suits them.
Neither of the contemporary advanced city design patterns mentioned are necessarily more requisite nor convenient than a hexagonal street layout as far as municipal budgets are concerned, even if it seems more convenient to laypeople familiar with grids and culs-de-sac. As the video author points out, hexagons are an extremely efficient shape, theoretically enabling greater density and simpler, safer intersections. If nothing else, contemporary lifestyles do value minimizing costs, and that aligns with efficiency. The video doesn't go into a lot of depth, but the main reasons hexagons haven't been adopted, according to the presenter, are more to do with stylistic preferences of developers.
I don't have much of an opinion on hexagons, but from the extremely cursory glance I have into the idea from this video, it does seem like an idea worth more investigation. I'm aware of little modern research on such designs at scale; this fact doesn't preclude research or discussion from happening. The kind of hopelessness embedded in "well, humans are just lazy and stupid, so it can't work" comments doesn't strike me as useful.
I just looked at an image of a stacked hexagon and it looks cool but the streets would be limited by the length of each hexagon’s side. Also it would lead to much more intersections which does not look like a great idea either. I didn’t watch the video so excuse me if they discussed these points.
The video talks about three-way intersections, which are apparently a lot safer than four-way intersections. This makes sense to me intuitively (fewer theoretical collision points). I wonder if it would be cheaper to operate traffic control devices on a system designed for such streets. Perhaps they could get away with being less complex if there are naturally fewer safety risks to the design. I'm not sure.
Do you feel it's important for streets to be long/straight? I wonder what traffic pattern flow would look like in a city based on hexagons but which has a variety of block sizes. Since hexagons are infinitely stackable, I don't see any reason why you can't just have a set of seven hexagons for a large public building. Some of the designs shown in the video seem to take this idea in stride.
Maybe I’m stupid but I don’t understand how hexagons are infinitely stackable while squares aren’t. To me this whole idea looks like a dream. Main cities are already built as is and nobody is going to rebuild them to have a hexagon layout. A lot of the cities not in America don’t even have this type of layout and are built chaotically.
As for streets being long and straight, that just seems more convenient to me to be able to orient yourself. For example if you turned to a wrong street while the layout of streets is perpendicular to each other, then you know that you can simply drive a bit further to the next intersection or two to correct your route, while in a hexagonal layout you would just get lost in all of the short streets with their unique names and their intersections.
You're not stupid. It's a concept we aren't used to. I had to draw out a bunch of hexagons to get a sense of why this might be beneficial and how its problems could be overcome. I'm skeptical on some of the finer details, but it seems like something that could potentially be useful to society.
Squares are also infinitely stackable. Hexagons just have a better ratio of interior area to frontage; on a structural level, they can potentially maximize the amount of space used productively by a city and minimize the amount of space used unproductively. Bees build hexagonal honeycombs for this reason, not square or circular ones. Squares, like hexagons, tessellate to form a grid; but hexagons are closer to circles than squares are. Circles are "efficient" insofar as their perimeter is as small as possible relative to the area within that perimeter. By contrast, a square "wastes" space in that it requires more of a perimeter ("envelope") for a particular usable land area. In the context of a city, this means more resources must be allocated to unproductive connections (roads) between productive nodes (blocks). Equilateral triangles tessellate into hexagons and are less efficient in and of themselves. Octagons and most other polygons do not tessellate. Hexagons, as the highest
n-sided polygon which tessellates, therefore resemble the "optimal" circle the most (reducing wasted space within a unit) while also forming a repeating pattern (reducing wasted space between units, enabling higher density).I am tempted to speculate that a hexagonal layout seems inconvenient merely because it isn't common, not because it is necessarily inconvenient. As a repeated, ordered layout, it has a predictable set of rules if you know what the pattern is. A straight line is obviously very simple, but hexagons also offer parallelizable routes, they're just intersected by blocks. It's like if you're walking in a square grid with a series of minor streets which are interrupted by offset blocks, which is common in many square grids anyway; Philadelphia's grid, supposedly neat and orderly, has elements of these interruptions in many streets (follow Cypress Street or Locust Street west to east, for example). If one were to grow up in a hexagonal city, perhaps one's understanding of navigability would be fundamentally different, and a square grid would be mystifying at first. Their worldview would be literally different; the way they understand layout might rest on things that us 90-degree'ers never even consider. I think these things are relative and not absolute. (In addition, the increasing prevalence of technology in our day-to-day navigation may, in the more distant future, change the way we understand navigability as a concept. Case in point: the Hexagonal Efficient Coordinate System. As computers get better at using hexagons, a navigational system based on them may become optimal.)
Presumably a hexagonal city would have a few arterial corridors much like a square grid, depending on traffic needs. But from a city planning perspective, completely straight roads maximize speed at the expense of pedestrian safety (straight roads universally increase speeds, and higher speeds universally kill more pedestrians), and the complexity of four-way intersections may reduce traffic throughput more than necessary, depending on lane orientation, pedestrian crossings, etc. In theory, apparently, a hexagonal system could maybe alleviate some of these practical issues. By reducing straightaways, drivers of multi-ton killing machines would, on average, kill fewer unsuspecting pedestrians. (Many systems currently in use on straightaways to protect pedestrians increase wait times so much for pedestrians and automobiles alike that pedestrians often cross without a signal, increasing their risk of injury. The longer the wait time, the more likely this is to happen, so if we want to maximize safety for pedestrians we probably have to decrease average travel speeds for automobiles. Pedestrians crossing streets where drivers are less likely to get up to high speeds, including less straight or more interrupted streets, may be less likely to be injured by automobiles.)
A hexagonal grid's shorter road length per given area reduces distance traveled for the average trip, which could reduce a lot of travel time. And while I'm not traffic expert or mathematician, it's maybe possible that three-way intersections also have a higher safe (for automobile passengers) maximum throughput than four-way intersections (even though four-way intersections possibly/probably have a higher maximum throughput in general). Four incoming roads mean there are twelve conceivable directions traffic could flow (three per incoming road: left, right, or straight), a ratio of 1:3. In contrast, three incoming roads means there are only six conceivable directions traffic could flow, a ratio of 1:2. In each case, 50% of turns are quite dangerous (left turns). But three-way intersections might have fewer theoretical dangerous collision points if someone runs a red light, because at a 120-degree angle they would possibly be more likely to cause a rear-end or a side-swipe than a straight-on t-bone. If "right-on-red" is permitted in a four-way intersection, then 6/12 (50%) of possible routes through the intersection can safely occur at once. If "right-on-red" is permitted in a three-way intersection, then depending on how you define a "relatively less safe collision point" (like a 90-degree t-bone) vs. a "relatively more safe collision point" (like a rear-ending, or a side-swipe), the number of possible routes through the intersection that can semi-safely occur at once is 3/6 (50%, if only right turns are permitted, and never left turns); or 4/6 (67%, if right turns are permitted, and left turns are permitted from one incoming street at a time, the other two streets yielding according to a traffic signal). With a 120 degree angle between the incoming lanes of traffic, rather than 90 degrees, it's almost like merging on the highway, which we don't consider unsafe. If integrated into a small traffic circle (roundabout), it's possible that safety and throughput would be further improved, perhaps more so than in a comparable four-way intersection. I'm not convinced any of these designs are necessarily safer or more efficient in practice, but maybe they could be if implemented thoughtfully. I tried to find some conclusive studies on this, but the data itself seems occasionally contradictory in places that study conclusions favoring four-way intersections tend to hinge on; and most data on three-way intersections is about 90-degree t-intersections, which are a fundamentally different design; so I'm not sure what to think. Some search terms pull up radically different results, I guess because they aren't technically all studying exactly the same thing. This 2022 study from Pravinvongvuth & Matarage seems to be one of the more useful ones for this particular discussion. Interestingly, safety in 120-degree intersections seem to have been studied more for pedestrian flow, as in Chen et al. 2019, which also seems to favor 120-degree intersections in reducing traffic flow conflicts. But it's unclear to me how universal this is.
I imagine the implementation would be city-dependent. Some geographies might be harder to develop this pattern in than others, as is the case with square grids. Intuitively, travel times can be minimized by connecting nodes by a straight edge. However, a square grid introduces very inefficient diagonal travel routes. If a city's geography is highly linear, it's important for its travel times along that linear axis to be minimized, which is perhaps most easily done with a square grid. However, this makes diagonal travel times comparatively slow; an inefficiency. Manhattan might be an example of a somewhat linear district. But in the bigger picture, much travel in our cities is not linear along the two axes of a square (or does not have to be). Manhattan is not just a line but rather a rectangle; and residents have travel needs to places that are not purely north-south or east-west, like going to JFK Airport if you live in Washington Heights. Even within Manhattan itself, orthogonally diagonal travel can be slow relative to linear travel. Speculating: given a plane containing blocks of equally distributed population density traveling in random directions, one could maybe optimize average travel times by using a hexagonal grid instead of a square grid, because a set of
nrandom node pairs on such a plane are more likely to include diagonal travel than purely linear travel. Thus a suitably large sample size of nodes would be more diagonal than not, and so would benefit from a layout which has relatively efficient diagonal travel. But I am not a mathematician. Perhaps there is something more complicated with travel times that I am not considering.Again, lots of speculation from a non-professional going on here.
That wasn't the suggestion of the video (have you watched it?) or anything I said. This is just an interesting design concept that we could use when building new cities or neighborhoods, or potentially for destroyed neighborhoods that are being redeveloped after a fire, flood, etc. Just because the status quo is what it is doesn't mean it is necessarily the best use of our extremely finite resources and money! The more inefficient our cities are, the more taxes we pay to maintain them (unnecessarily). It is always useful to think about alternatives.
I do appreciate the comments, because I'm not wedded to this concept by any means and am interested in other perspectives, but urban planning is the kind of field most susceptible to inertia and which benefits the most from ideas challenging the status quo. The hexagonal grids envisioned in the video do look a little dreamlike, but I don't see a great reason why they're automatically something to throw in the wastebin. Our cities today look absolutely nothing like the cities of 1000 or even 200 years ago; perhaps our cities 200 years from now will take after the humble bumblebee. Who's to say?