It's time cyclists discovered OpenStreetMaps. The cyclist community, as far as I can see (it's just a mode of transportation for me, so I'm not part of it), has been doing an excellent job of making sure cycling routes are properly mapped there. Doesn't solve the Street View problem, but you're really missing out if you haven't tried it for navigation.
Even in the Netherlands, it's often better than Google Maps. And it can serve as the basis for apps building on top of it like https://cycle.travel/.
cycle.travel's my project - thanks for the mention! Always happy to hear suggestions and ideas.
It uses OpenStreetMap data plus a heavily customised instance of OSRM (Open Source Routing Machine, a fast routeplanning engine). The routing "profile", which calculates the weighting, is a few thousand lines of code alone, never mind various other stuff like custom elevation calculations.
It's a great project to work on, and the best reward of all is when people plan multi-day tours with it and say how much they've enjoyed the route.
Bookmarked! While it seemed unable to locate my home address for some reason, when I manually dragged the pin it planned a route from my work to my home nearly identical to the one it took me several weeks of harrowing journeys with Google Maps and looking at PDFs of bike routes to find. This will DEFINITELY come in handy in the future!
This is great - bookmarked! I see that while it does try to avoid busy roads, it often takes the road that's right next to the motorway (even when not strictly necessary - probably because it's shorter and has little traffic).
It'd be better to prefer roads that are further from busy roads. I understand that's very hand-wavey and probably super tricky to implement.
It would! As you say, roads (or cycleways) beside motorways are usually short, have little traffic, gentle gradients and often a prized position in a valley rather than lumbering up a hillside. Ideal cycle routes... apart from the motorway a few metres (/feet) away.
I've got a page of notes on possible ways of fixing this - it is a hard problem. One possibility I'm toying with is a massive grid of "scenic value" for (say) each 0.01° x 0.01° square, so I could downrate the surroundings of motorways and uprate routes through (say) national parks.
Calculating "scenic value" sounds like a very good idea indeed! Plus points for points of interest and greenery, minus points for busy roads. Again, I understand it's not easy to implement :)
It generates a ~10km route in my Canadian city that starts off in meters, and switches to miles. I know we're somewhat used to flipping between metric and imperial, but it's not a common thing to do mid-trip.
So I’m genuinely interested - what would you expect? I was born in Britain in the mid-70s and miles and metres, though an unconventional combination, is what I grew up with. But I realise not everywhere has the same weird combination of metric and imperial as we do. What would make sense to you from a North American viewpoint?
Edit: Thanks everyone for the feedback! Just pushed a quick fix:
- USA now defaults to miles & feet
- Canada and Mexico now default to km & metres
- Britain defaults to miles & metres (as before, yes we're weird)
- mainland Europe, Australia and NZ default to km & metres (as before)
I'm from the US, and I would never expect "feet" in a road sign or on a bike. For distances below a mile I'd expect "½ mile" or "¼ mile"—this is how they are almost universally marked on road signs across the country. For extremely short road distances I've occasionally seen "⅛ mile". For road distances any shorter than that, which in my head I'm thinking of as "very short long distances", at that point I might expect—and I feel a little guilty even typing this out loud—yards. (Or blocks, where 1 block = ⅛ mile, but that's a peculiarly Chicago measurement, I think.)
Feet do show up on signs, but always in large round multiples of 100 feet. They tend to only show up on low-speed roads though -- signs like 'blind driveway 500ft'. On highways, distance signage is almost exclusively in miles and fractions thereof.
As an American, the mix of meters and miles really threw me off. When it said "40m" and "5mi" I thought it was switching between estimated travel time and distance. For me, it would be better if it was consistent with the measurement system (using miles and feet, OR meters and km).
I immediately recognized metres + miles (for a road distance) as British, I was also born there.
The USA should be feet and miles.
Canada should be metres and kilometres.
But you did say North America, it seems these might use miles per hour for road speeds, so they might not want metres and kilometres. I don't know if they want feet, yards or metres though.
Antigua, Bahamas, Barbuda, British Virgin Islands, Grenada, Puerto Rico, St Kitts and Nevis, St Lucia, St Vincent and the Grenadines, or the US Virgin Islands.
In canada we use feet and inches for building dimensions, but not for distance travelled. so we've still got the weird metric/english combo, but not quite as weird as miles and metres.
cycle.travel does not work on iOS. Loaded page, green popup, Starting from: cycling to: I filled these out and pressed Find Route, nothing happens. Tried selecting Circular route, press Find Route, nothing happens
Curious - works for me but would like to debug! Could you ping me the start/end points you're entering? (info@cycle.travel if you don't want to post them here.)
Works for me on iOS 13.4. Even have 1Blocker installed; didn't have to disable.
Edit: This looks pretty great, but I should probably point out that it turned my 5.5 mile commute into a 8.5 mile one. Realize it's probably trying to avoid larger streets; Hope this is or becomes configurable.
Glad you like it! It does optimise for low (motor) traffic routes. But you'll see that there's a little "Any road/Paved only" toggle - currently the latter just filters out unpaved roads/paths, but my medium-term ambition is for it to be a bit more blasé on city streets too.
It looks like OSM has two builtin bike routing algorithms you can use, and at least for a few routes I tried they both seem significantly worse than what Google Maps does (Big fat disclaimer: I work at Google, although not on anything related to this).
The OSM directions for biking to my office involve going through a decommissioned oil field (fenced off, so the road isn't actually passable) and a swamp. What's weird is that it looks like OSM knows the road is closed but still tries to route me through it. Google's directions for the same destination aren't perfect either but at least they'd get me to my destination (they include a short jaunt down an eroding hillside where I have to walk my bike because the sand is so loose).
That cycle.travel site looks really promising. They give the same route as Google Maps, but do a much better job explaining that part of it is an unpaved dirt trail.
OSM itself is really just the base map. A huge, editable database of map data. Anyone can build apps that make use of that data - the route planner you used is just one of many.
The routing errors you describe could be issues with the particular route planner you used, or they could be issues with the OSM data itself.
Some parts of the world are indeed better mapped than others (this is also true of Google Maps), but the great thing about OSM is if you find an error you can fix it yourself in seconds.
Just click that “edit” button right there in the OSM interface...
Yeah, when I was referring to OSM I meant that map data they collect, which has excellent bicycle data. That can then be used by specialised projects for cycling-specific interfaces.
I'd have to toss my bike over a barbed wire fence, or maybe cut the lock. The service road winds between a bunch of natural gas storage tanks so I don't think they'd willingly let me in. There's a "≠" symbol where the service road is fenced off, which I assume means impassible but "≠" isn't on the map legend so I'm not sure. When I hit the cycle layer, the ≠ symbol goes away
≠ represents a gate [1], which the routing engine probably assumes you can go through; if it's not possible to open the gate, it should have an access=private tag added to it to indicate that it's not passable.
If you click on the mouse cursor button on the right and then click on the symbol, it will show you the details for that object.
I have to agree, this instance that "Google" has to do better is silly. There are alternatives that are dedicated to cycling and mapping of cycling routes. Why not use one of them?
Nobody needs a single entity corporation to do everything for them, it's really time people see that there is shopping experience besides Amazon and search/map experiences besides Google.
Why shouldn't Google do it? We should want as many people as possible to cycle. It's healthier, reduces traffic, no pollution.
Google maps is the default option for most people. Why would we want to make it harder for people to start cycling by forcing them to hunt down some specialist app that they don't even know exists? Why not just have the information in the app they're already using?
Because as “technology”* becomes infrastructure it’s better off that it transition (via growth) into public hands. And OSM (and Wikipedia and..) demonstrate that “public hands” doesn’t automatically mean “government ownership” especially with transnational resources.
* in the general sense, not just today’s sense, like roads, bridges, power, water etc.
As a cyclist, is there an app I can install that simply has “start recording” and “stop recording” and lets me trim my off the start and end, so that I can contribute my cycling path data to OSM without having to curate it further?
The set of people who would happily upload cycling data minus their precise start/endpoint is probably 1000x larger than the set of people who enjoy map curation in their free time, but I’m not sure what solution I can present to bicyclist friends that would let them contribute.
Just having a trace of your route probably isn't very useful in a lot of cases, since it's going to be mostly on roads that are already mapped. Instead, you'd probably want to map the actual paths, in which case you can view your GPS trace in the editor to map from without having to share it publicly.
The Strava heatmap is also licensed for usage in OSM, and in a lot of areas it has pretty good coverage of bike paths already (https://wiki.openstreetmap.org/wiki/Strava).
I don't think there's any exact numbers. You can search for changesets that have Strava listed as a source [1], but that doesn't mean the Strava data was actually used.
A lot of cycling routes are obscure and labyrinthine, meaning you have to constantly check your phone to ensure you aren't lost. While riding, and in direct daylight I have to stop to be able to read the map. It ends up taking a lot of time. I wish cities were more clear about signage. I experienced obscure and labyrinthine bike routes in my own city, and I usually bring a bike with me when travelling to other cities and have experienced the same thing pretty much everywhere.
The thing I would much prefer is clear signage. The lack of it is common.
Totally! All too often cars have a direct route from point A to point B while cyclists have to wind through complicated, low-speed side roads with little or no signage to help.
That said, I think Portland has done a great job with the signage part of this problem. Signs like [1] are present at just about every intersection of bike routes, giving directions to neighborhoods and landmarks. And when a bike route dog-legs, directional "sharrow" symbols [2] are clearly visible on the road surface.
My favorite biking accessory is my Quadlock - https://www.quadlockcase.com/. Having my phone securely mounted on my handlebars is a game changer. If I want maps, I have those up. On longer rides, I have my phone playing music directly from the phone speaker which I actually like more than using headphones. I also have my heart rate monitor connected to my phone so I can see those stats if I want them. My bike feels like a Tesla with a touchscreen on it.
A battery pack case + Quad-lock is what I used to use. I got a Garmin for the battery life. But I didn't find that useful for navigation - slow and really hard to set destination and doesn't always pick the safest bike path (Google isn't great either but in my area it consistently chooses inner bike path over path that run through multi-lane roads).
I was going to say. This is a GIS problem, not a Google problem. OpenStreetMaps is a phenomenal way to approach this problem. Distilled down, you care about different properties of your spatial datasets. Cars don't so much care about topography, but you do. That data is all there.
I'm also Dutch, and even for cars Google Maps isn't the best. Traffic info maybe, but routing? Here Maps does both, with better time estimates in my experience.
Not to be pedantic, but OSM is just data, you're asking whether there exists software that can do this. (You might for example ask in the OsmAnd Telegram chat whether this can be done with OsmAnd.) I don't know if there is any that sticks to only cycle paths, but there is definitely software that'll avoid car-only roads like highways, try to avoid big roads where reasonable, and prefers cycle paths where available, such as OsmAnd (you can also configure routing preferences like shortest way, prefer byways, configuring your cycling speed, using elevation data...). For turn by turn instructions, that seems to be pretty standard and OsmAnd can do this too. Have you tried any software that uses OSM data?
Interesting article. One challenge though is that there are a lot more gray areas in cycling than in driving. Driving is black and white - you want the fastest route thats legally possible. With cycling, everyone's tolerance for what is a good cycling road is different, and you are constantly making trade offs between road quality and scenery, and distance.
The cycling directions in my town, for example, are great in the sense that they take you on only bike paths with little cars. However, they are often 25% longer than just taking the roads which also have bike lanes. I am comfortable biking on the roads so I prefer the roads, but others might not be. Which set of directions do you present? I guess you could have options for both like "most scenic" and "most direct". However, there will always be people complaining about routes being too roundabout or not safe enough because of preference differences.
I think with cycling sometimes its best to just use the bike directions as a starting off point and then explore from there. Hopefully you can be in a place for long enough to discover the routes that fit you best and stick to them. For me, thats actually a fun part of the activity.
Most map apps allow some choices like 'avoid tolls', or 'avoid highways'. Biking should/could be similar. Prefer bike paths vs bike lanes. To be fair, it is hard for the Google Street view car to go down the bike lanes. Going to need to rig up a bike with the camera rig. They've done it for hiking trails, so why not?
Those choices are insanely expensive for Google to provide.
Every unique configuration of options has to be hosted by hundreds of servers in each region. The cost of running the routing service is proportional to 2 ^ (the number of options). It's because everything has to be precomputed for fast lookups, and the lookup tables for one set of options aren't valid for another. See [1].
This is why Google hasn't added any more options in ~a decade. Even then, some combinations of options use various approximations and in some cases don't return the mathematically best result, or take longer to do so.
That paper is old and definitely not what Google Maps actually uses. Techniques which people actually use (mostly Contraction hierarchies and CRP) are much more compact than what you describe. Routing isn't just a table lookup.
Google Maps' offline maps also include routing information. The downloads are pretty small, which definitely excludes the more space-intensive algorithms.
To be fair, contraction hierarchies are inflexible for exactly the reasons OP described. CRP (aka multi-level dijkstra) is more flexible, but there's always a trade off between query speed, pre-process/weight-update speed, and route quality. These services are very expensive to run at global scale if you need multi-modal or live congestion support.
Customizable Contraction Hierarchies go a long way towards solving that, they're quite similar to CRP with regards to both customization and query time. Also, route quality isn't part of the trade-off: all of these methods are exact.
I would argue that updating the routing algorithm's data structure isn't the expensive part of reacting to live traffic. Getting and processing the raw data sounds a lot harder than re-running the customization phase of routing preprocessing.
I'd be very concerned if download size wouldn't scale with the size of the map, so I'm not quite sure what your point is. Nobody uses lookup tables for routing (that would require quadratic space to store). If you really want to use something that could be described as lookup tables (even though "set intersection" would be more accurate), have a look at hub labeling. But that still requires a lot of space and isn't particularly suitable for traffic updates, so it's clearly not what any of the maps services use, online or offline.
”It's because everything has to be precomputed for fast lookups”
I don’t see how that can be true. With, conservatively, 50 million destinations, that would mean 25E14 different precomputed routes (edit, that’s exactly what the paper you reference argues: An extreme way to accelerate queries is to precompute all shortest path distances d(s, t). However, this is impractical for large graphs since it requires quadratic space and preprocessing time.)
Also, the average GPS without any internet connectivity can compute routes. Surely, those don’t store those precomputed routes.
What likely is true is that it may be a bit harder to do a bicycle route planner. Slightly simplifying, a planner for cars, given two places A and B relatively far apart, ‘just’ needs to find the best route from A to any on-ramp to a highway, from any highway to B, and a route only on highways between the on- and off-ramps.
On the other hand, bike routes that have to be planned typically will be a bit shorter.
Together, that means it isn’t that hard to compute routes on the fly. https://routeplanner.fietsersbond.nl/ does exactly that, giving you options such as “shortest route”, “fewest interruptions”, “racing”, “fewer cars”, or “recreational”.
It's kind of true. Routing algorithms like Contraction Hierarchies and Transit Node Routing don't precompute every possible A-B pair, but they do store so many "shortcuts" (i.e. the best route between common As and Bs) that a route lookup will quickly converge on one of these precomputed routes for all but its first and last sections.
It's similar to the way that we intuitively navigate by car - if you drive from (say) London to Glasgow, you don't bother weighing up the many options: rather, you take whatever local roads get you to the M1 fastest, then follow M1, M6, M74, and then whatever local roads get you to your exact destination.
Would it be cheaper to download the bike route data to the device and simply do the processing there?
Generally bicyclists have a fixed starting point so you could cache the results and improve them over time. (Obviously cross-country would require further work to support efficiently.)
I’m all for “anything can be done in the cloud” but if the cloud can’t efficiently handle this, I still have a couple of supercomputers laying around that I’d be happy to let puzzle over my bicycling needs for a few minutes (or overnight, or for a few weeks when charging) if that helped me find a more scenic route someday that isn’t much slower than the direct one.
I'd speculate that most of that can be attributed to the TomTom's status a local device. I don't think the trouble is doing this for one person, but rather doing it for thousands to millions of people simultaneously. Also, the TomTom is probably operating on a pretty old set of static maps, rather than the constant updates that a service like Google Maps gets
>Those choices are insanely expensive for Google to provide.
That's a weak hill to die on with the money G makes. That's just the cost of doing business. People using maps need choices. If you as a map app choose not to provide an option, that's just an area a competitor can one up you. Once they do, you have to play catch up. That could be advantageous as you see what the competitor did that does/does not work allowing you to shortcut some of the R&D.
It's not "doing business" if people using maps don't provide enough revenue back to the company either directly or through advertising. The reason Google makes so much money is that they are careful to implement features that are revenue positive.
Google isn't bottlenecked by money. Computers don't run on money. They are bottlenecked by getting datacenters online and staff ability to manage the complexity to keep systems running.
I don't really buy this argument: each pre-computed route could be tagged `{toll: false, highway: true, elevation: 15, scenic: 0.9}` and such, making it almost trivial to filter given options.
The expensive part is not the filtering. Actually you can do hundred of thousands of road segments per second (maybe 10x less on a phone). The expensive part is the visited nodes from Dijkstra vs. the much less nodes from the query of a precomputed request. The speed difference can be 50x or 200x, depending on the route length.
I don't think you understand how these routing algorithms work. Contraction Hierarchies, Customizable Route Planning, Transit Node Routing, etc are all guaranteed to give you the same result as a Dijkstra search. They just exploit properties of road networks to do so much faster. The secret sauce of routing is not in the algorithms (those are publicly known), it's computing the edge weights (including things like turn costs). Effectively, the edge weight has to be the desirability of driving/riding/walking along that particular stretch of road. A big factor of this is obviously travel time. This whole discussion is centered around the observation that whatever metric Google uses for bicycle routing, it doesn't result in routes that users perceive as good.
I'm not sure I fully understand this. Is the speed difference caused by the options limiting the use of pre-computed nodes? As in, looking for guarded bike roads limits the search for pre-computed nodes' lengths and the search has to be deeper as a result of thay.
Should/could be similar? It "is" similar! There is software that allows to configure your cycling speed as well as such preferences, one example I know of is OsmAnd.
It would be great if there was a more objective rating system to refer to (maybe something like this exists and I'm not aware of it?). Like an A bike rating would be a truly protected bike lane, physically separated from moving traffic as well as parked cars and loading curbs and anything else. B might be a dedicated bike lane, but not physically separated from road or parked cars so you might get a door opened on you.
Then you could look up map directions and maybe choose different routes based on the time estimate plus a breakdown of what percentage of the trip will be in different conditions, and you can choose your preferred route based on your safety tolerance.
As somewhat of an aside, I find the way you've explained your trade-offs here strange - as a matter of tolerance, and comfort being around cars vs desire for scenery. I've seen a lot of stories of very experienced, comfortable cyclists killed by cars every year. IMO this is very much not a learning curve or experience issue, it's an issue that biking on busy roads designed for cars and not bikes is genuinely dangerous in a way that is largely out of your control as a cyclist.
Luck certainly plays a factor but experience cycling on the roads is key.
When I cycle on roads I am hyper aware of my surroundings. I am scanning all around me, including behind me every few seconds. I’m also specifically looking at people’s faces and assume they don’t see me until we make eye contact. I also take space aggressively - if there is no bike lane I will bike in the middle of the car lane until I see a safe place for the car behind me to pass, and only then move over. This might piss off the cars behind me, but as long as they see me, its OK with me.
All these techniques make sure that I am seen by other cars and nothing surprising happens. This kind of riding takes years to build up experience. Anectodal, but through my 25 years of heavy commuting on a bicycle on roads shared with cars I’ve never had an accident - but I’ve certainly had several while driving a car. Ive also had several close calls on bike trails because those tend to be filled with pedestrians who aren't looking at all and don't follow any rules.
For sure I could go out biking tomorrow and get killed but freak accidents happen in all areas of life. Personally I don’t feel in danger on the roads - but everyone’s tolerance and experience is different and that’s one of the cruxes of giving good bike directions.
How much car exhaust do I want to sniff? How much do I want to have to fight for space on a 'cycle lane' that's just a bunch of road markings directing me right down the middle of a lane of cars? What's the shade situation like? How well-maintained is the road? How sketchy are the neighborhoods I'm going through? Am I in a hurry, and how much do I wanna trade these other desires off for The Shortest Possible Route? Am I taking my time and feel like exploring?
At least I currently live in a very flat place so I don't need to worry about any hills to either detour around or sssssllllooooowwwwlllllyyyyy grind my way up.
Your preference may even change depending on where you're going. For my daily commuting and basic errands, I pretty much always take the most direct route. But then, I don't need an app for that, since it's the same every day. Even for recreational riding, I have a bunch of routes that I know well, and really only need to consult a map if I'm somewhere unfamiliar. And then it's usually not 'tell me how to get home' but 'get back to somewhere I know'. Usually a minute of scrolling around on the map will do that.
I would love a scenic option in driving instructions, much less biking instructions, but I think there’s a nuance that’s missed by that phrasing.
Imagine a slider between “Prefer protected bikeways” and “Prefer fastest routes”. You could show in UI the trip time for each extreme and a ‘middle’ choice.
Paired with a slider that says “Avoid uphill climbs” to “Prefer uphill climbs” and some checkboxes for scenic routes and paved-or-not and “fragile goods”/trailer-towing/etc, you’d have something pretty amazing.
That's my biggest problem with Google maps for cycling in London. They somehow always want to send me via Regent's Canal even if it's farther. I get that there are no cars and it's theoretically a nice route, but way too crowded to cycle during the day.
The thing about bike commuting for a couple decades is you quickly learn that about half of the streets and and bike lanes are complete no go zone death traps, and there is no good way to really build this knowledge aside from just learning where you're going to get turned into repeatedly.
My city is super progressive and bike friendly, but even then it's effectively suicide to try and use certain bike lanes and everyone just has this communal knowledge.
a) capture routes of bike trips taken while using nav apps
b) send camera-laden cars around, the same way these maps were originally built, and this time, measure bike trips
c) ask bicyclists
If everyone really has this communal knowledge, as you say, then it can be gathered.
> capture routes of bike trips taken while using nav apps
Still fails to capture individual preferences which are often complete opposites, sometimes even for the same person on different occasions.
When I'm riding fast I greatly prefer proximity to cars over proximity to pedestrians or slow cyclists, it's much safer because even the worst drivers are more predictable than average pedestrians, and the speed difference between fast cycling and urban driving is quite small anyways. In terms of injuries it doesn't matter much who takes you down. When I'm riding slow those preferences (and the safety considerations they are based on) are completely inverted.
Dunno about no go zones, but the bike lane next to my old place was pretty horrendous. It went down a fairly steep slope next to a four lane road, about halfway down the hill the lane cuts across into the middle of the street about 100m up from a 4 way intersection with a highway. You have the cross the highway in the very center of the intersection, between all the cars, then on the other side, the bike lane cuts back across traffic, then vanishes for a block or so before starting again on the sidewalk.
The first time I rode my bike down there I terrified. You don't have much time to react, the hill creates a blind spot at the point where you have to cross traffic and because the road is used by people trying to beat traffic during rush hour, drivers speed constantly and don't really pay attention. Just on foot i've almost been hit by cars a few times at the top of the hill. Definitely not my favourite route.
> even then it's effectively suicide to try and use certain bike lanes and everyone just has this communal knowledge.
As a lifelong city biker, I am curious what exactly you are talking about. Certainly, there are some roads where you're more likely to be hooked by someone turning on right (Market St in SF comes to mind), but not a "no go zone"?
This is interesting. I know of places where it's dangerous to ride in my city (for one of two reasons: 1. more accidents happen there statistically or 2. narrow shoulder / no bike lane so it's unsafe unless you occupy a lane of traffic) - but what do you consider "death traps"?
One thing I've heard of is "The Right Hook". The bike lane is on the right edge of the road. When a car is making a right turn, they don't look behind them, make a right turn, and a cyclist crashes into them. Seattle has a similar problem with some 1 way streets with a bike lane on the left side. At least 1 person has died.
This is a fundamental traffic design failure. Prior to the existence of bike lanes, motor vehicles never had to yield to something behind them (perhaps 30 or 50 meters behind them) while traversing an intersection.
In a busy intersection where one also has to attend to pedestrians, other cars, etc., moving in different directions, trying to track what's going on behind you can be almost impossible (and thus quite dangerous for all).
My "solution" to this problem is to carefully but fully occupy the bike lane before making such a turn. Yeah, it's illegal, and draws lots of swearing (and worse) from bicyclists. But there's just no other good way to keep everyone safe.
> My "solution" to this problem is to carefully but fully occupy the bike lane before making such a turn. Yeah, it's illegal, and draws lots of swearing (and worse) from bicyclists. But there's just no other good way to keep everyone safe.
Funnily enough, that's exactly what you're supposed to do and is required by law in every state i've lived in.
First off, thank you for doing what you're doing as a driver to prevent this kind of crash. You're right, cyclists are blithering idiots about drivers doing this, but it protects them even if they don't understand that.
As a cyclist (about 15,000 total miles, half touring, half commuting split between Boston and Cleveland), my solution to the problem is to not design the infrastructure in ways that sets cyclists up for the crash.
It is absolutely woefully bad design that cycle lanes on the right continue all the way to the intersection. A solution that works for experienced cyclists to to end the bike lane some distance before the intersection. This forces a merge in a zone where there's no risk of a right hook. I float this with the full understanding that it only works for cyclists who are capable of making the merge at around 20mph/30kmph. And/or where the topography is favorable.
No, you don't have to be able to sustain that speed, but being able to do it in short bursts greatly expands your choices when riding in urban traffic.
My experience has been that nobody minds much at that speed, especially if you merge back into the "bike lane" while you're still in the intersection but after you're past the lanes people would be turning right into. Turning traffic has to slow anyway to make the turn, after all.
This is a specific case of a general rule I try to apply when riding: Make people choose to hit you; don't let them do it accidentally.
I would add that this is really only a valid strategy in urban traffic where even the major arterials only flow under 30mph because the density of turns and stop lights prevents higher speeds. Major arteries in the suburbs or around malls are a nightmare for riding because traffic generally flows a lot faster. Some of the most hair-raising riding I've done was near a mall while bike touring. It was basically unavoidable, and also absolutely miserable.
> A solution that works for experienced cyclists to to end the bike lane some distance before the intersection. This forces a merge in a zone where there's no risk of a right hook.
Unless the road has sufficient room for a bike lane of a minimum width of 5 feet, then you end up with a lot of close passes from cars in the right lane and also end up riding in the door zone of parked cars to your right.
In those cases, it makes more sense to just ride in the center of the right-most lane available for traffic and get more room from cars passing you. This would work for streets that have multiple lanes for same direction traffic.
> Major arteries in the suburbs or around malls are a nightmare for riding because traffic generally flows a lot faster. Some of the most hair-raising riding I've done was near a mall while bike touring. It was basically unavoidable, and also absolutely miserable.
It really depends on how much traffic is present. I've ridden on roads in the center of the right lane with speed limits ranging from 35 to 45 mph without much of an issue. Drivers almost always change lanes to pass me or slow down behind me before passing.
> Unless the road has sufficient room for a bike lane of a minimum width of 5 feet, then you end up with a lot of close passes from cars in the right lane and also end up riding in the door zone of parked cars to your right.
Thank you for clarifying. Yes, cyclists riding in a regular traffic lane should absolutely take the center of the lane for all of the reasons you've listed. Again, make people choose to hit you; don't let them do it by accident.
> This is a fundamental traffic design failure. Prior to the existence of bike lanes, motor vehicles never had to yield to something behind them (perhaps 30 or 50 meters behind them) while traversing an intersection.
This is true. The closest thing thing to this is checking for pedestrians, but even the fastest pedestrians aren't moving more than 7 to 8 mph and will almost always check for vehicles prior to entering an intersection. A cyclist could be moving anywhere from 10 to 20 mph and could be quite far from the intersection and still end up in a crash with a turning vehicle.
> My "solution" to this problem is to carefully but fully occupy the bike lane before making such a turn. Yeah, it's illegal
In most states in the US, it's not illegal. The law states that right turns must be made as close as practicable to the right hand edge or curb.
> My "solution" to this problem is to carefully but fully occupy the bike lane before making such a turn. Yeah, it's illegal, and draws lots of swearing (and worse) from bicyclists. But there's just no other good way to keep everyone safe.
Which state/country are you in? This is actually required in California, and I'm always annoyed at the majority of drivers who simply DO NOT come and occupy the bike lane intentionally, making it confusing for me (as a bicyclist) whether they are going to wait for me or not. (coupled with how a lot of drivers just don't use turn signals accurately).
> My "solution" to this problem is to carefully but fully occupy the bike lane before making such a turn.
In California, that's not illegal, it's the law. It's on the driver's test. Kind of counterintuitive at first, but seems much safer and also more consistent with the way turns are handled with any normal lane.
At least in California it's illegal not to fully occupy the bike lane before turning right. But in reality most cars don't do it and most cyclists would get mad if they did.
EDIT: Okay, it's a surface-grade light rail train. In the cases I'm familiar with, there's always a red light forbidding crossing through the path of such a train. So, you're not yielding to the train, you're obeying a signal that's in front of you.
Speaking as a cyclist, it's extremely uncommon that you have to put yourself in a position that you can be right hooked. Cyclists need to educate themselves about road safety and take a share of the responsibility for their own safety.
What really pisses me off about this is that the only time people discuss the safety of cyclists is in the wake of a fatal crash. If you dare to suggest that cyclists often have choices that can prevent the crash, you get called out for victim blaming.
It's bullshit to shut down discussion of things people can do to protect themselves because of some misguided respect for the dead. For goodness' sake talk about it so people on the streets are armed with the knowledge that will save their lives so we aren't having the same discussion again and again and again.
It doesn't matter what the law says, it doesn't matter who had the right to occupy that bit of the road, it matters that cyclists keep dying because they're the more vulnerable road user.
It's naive to think that we're going to train every driver to do the safe thing. It's far more realistic to educate the people with the most to lose (i.e. the cyclists) to do the safe thing regardless of whether the drivers around them are aware of them or not.
> it's extremely uncommon that you have to put yourself in a position that you can be right hooked
I'm not so sure. I think there are definitely things to do that can make you avoid the vast majority of these incidents (don't bike to the right of a car when an intersection is coming up if you can avoid, obviously look at the car's turn signals before passing them on the right, pass on the left when you think someone will turn right, etc.), but there are plenty of instances where it is not avoidable and even experienced cyclists can get right-hooked (like biking on a bike lane on a heavily trafficked road with many side streets to turn off of).
> It's naive to think that we're going to train every driver to do the safe thing.
I think it's not an either-or. I don't think that we have to throw in the towel on every effort to change behavior - people have been convinced to stop littering to a massive extent, checking your right mirror seems minor in comparison.
> like biking on a bike lane on a heavily trafficked road with many side streets to turn off of
With the caveat that my experience is not your experience, I respectfully submit that the bike lane is poorly designed in this case and that it shouldn't be used to travel straight through the intersection in this case. I signal a lane-change to the left before the intersection and go straight through in the regular traffic lane, moving right ASAP after the intersection to not obstruct traffic.
My riding is mostly dense urban where traffic is moving slowly enough to do this if I put some hustle into it.
> I think it's not an either-or. I don't think that we have to throw in the towel on every effort to change behavior.
I am absolutely in agreement that it's not either-or, but if cyclists are doing something that's (almost) always safe, they're a lot less likely to be killed by the untrained drivers, of which there will be many for the foreseeable future.
What I do doesn't work for cyclists puttering along at 10mph (or grinding up a hill); the speed difference is too great to safely merge left. Training the drivers greatly increases the safety level for such cyclists on roads with cycle lanes between the regular lanes and the curb.
This is fair, and what you do is essentially the same as what I do and I am also mostly biking in urban cores.
That said, I think that we can't just place the onus on bikers to both ignore the drawn on bike lane lines and also to bike considerably faster than they are comfortable without trying to also change driver behavior. I think this mostly agrees with your last point.
Right hooks mostly happen when a cyclist is overtaking a car on the right hand side. This entire class of collision can be avoided by not passing on the right. If you see a car slowing down mysteriously, even if it is not signalling, assume it will try to kill you and keep your distance.
(I've ridden about 800 miles by bicycle so far this year, and >4000 total since I started using a GPS. No right hooks or right-hook close calls.)
Clarifying question: One ways with an opposing bike lane on the left side, or one ways with a same-direction lane on the left side?
Boston (and the surrounding communities) are doing an increasingly large amount of the former. I no longer live there and never used them much, but I'd be keen to hear about the hazards of that design.
Particularly dangerous to slow riders who can't modulate speed to time their crossing to a gap between cars (and they won't even try because slow feels safe)
The biggest problem is usually avenues (four lane roads, two in each direction) with high speed traffic. Around here that means lots of people turning left and right and cars average about 40mph, many do 60mph.
Unprotected bike lanes on these roads exist but over the past decade I've personally watched two people die on bikes from right turns and heard of dozens of collisions.
> In short, it treats the cycle as a car, but with access to less information.
This is just not even remotely true.
I cycle all the time in Manhattan and Brooklyn and for the same destination, the car and cycling routes are drastically different.
Cycling routes give tremendous preference for roads with actual cycling lanes, and do their best to avoid main arteries to give you a safer side street instead.
You can verify this yourself in Google Maps by turning on the "Bicycling" layer to see the green-marked bicycle routes, picking random intersections to get directions between, and observe how it tries to route you along bike lanes. (Of course, Manhattan streets are tricky, so it's not the case 100% of the time because giving too much preference to bike lines can result in a much longer trip.)
Quick searching also reveals that Google absolutely uses the elevation data to try to route riders along flat roads, that uphill shows a longer estimated time, and you can even see an elevation map of the route.
So as far as I can tell, this article is just 100% factually wrong, based on an entirely false premise.
Is Google Maps perfect for cyclists? Of course not (e.g. it has no idea cobblestone streets are something to avoid), but it's not perfect for cars either (thinks I ought to drive down a pedestrian staircase in LA).
But there's clear evidence that Google Maps "discovered" cyclists sometime around 2014, if not before. It's been a while. (My biggest wish is for it to integrate bike sharing into directions.)
Portugal doesn't have any cycling info. The routes in Spain and France tend to go down things that are marginally identifiable as roads or cycling paths, and more likely to be hiking paths. Definitely the sort of thing that I'd not take a road bike down, but a MTB would be ok.
Yes. I wonder to what extent that has to do with countries/municipalities providing data though?
I know there are huge variances in the quality of government-provided data, as well as whether governments provide them at all.
And that while Google Maps has Street View as a ground truth in some parts of the world, it doesn't everywhere, and especially often where there are bike paths that aren't next to a road for cars.
So without knowing more in any particular circumstance, the government may be just as likely to blame as Google.
> My biggest wish is for it to integrate bike sharing into directions.
This is why I always use Citymapper to find the nearest Citibike racks to my start and end points, then switch over to Google Maps to generate my route. It does a better job than Citymapper at actual bike directions, IMO.
You can always tell what features the creator of a product use themselves!
Microsoft long ago (10 years?) added a great feature for bus routes on Bing Maps: "Previous stop is ___" / "If you reach ___, you've gone too far". That's a brilliant idea, and it tells me that the people at Microsoft who built this actually take the bus themselves.
It's pretty clear to me that everyone at Google drives.
That one was a great feature! I'm really hoping someone will figure out landmarks for mapping directions at some point. Or has someone done that and I haven't noticed? :) There's so much data about buildings now, it wouldn't be that much of a stretch to say something like "head towards the Dominion building", or "once you've passed the school on your right, you are almost there". It would really help people who have trouble orienting themselves with maps, it would help in places with less obvious street signs, and it would be a hundred times more useful than that very cool and very unnecessary AR navigation feature Google made because they needed to work around the broken compasses in their phones.
I'm amazed nobody has mentioned Komoot https://komoot.com/ yet. They have an excellent route planning tool for cycling, at least in Europe. I can't vouch for other areas of the world, though.
The important point to note is the data they use: it is OpenStreetMap. It is such an incredible project and there are many great navigation apps. And yes, one of them is Komoot. But my opinion in this regard might be biased as I'm one of the developers of the underlying open source routing engine they use.
Strava just improved their mapping recently on the mobile.
You draw with your finger a shape, and in seconds(it is impressively fast) it gives you a route to follow. Based on their heatmap probably.
I then still have to import it in Komoot before using it, to filter out unwanted surfaces or way types.
I tried it for a few rides already and the route selection was great.
Komoot's routes are great for me as well. What I really dislike about it is its process for "stopping" navigation -- it seems to require about 7 taps to bypass all the "social" features (share my route, tag friends you cycled with, add photos and a narrative about how beautiful the route is, etc.) that I'll never use because I just want a goddamned navigation app. But I get it, gotta be able to tell the VCs it's a data play and put food on the table somehow.
(Edit: to be fair I just did a count -- it was 1 tap + long hold + 4 regular taps, though each is on a different part of the screen so you can't just button mash your way through. So I exaggerated a bit, it's more like 5.5 taps instead of 7.)
Hmm, I followed the link and just tried it for a 100-120 km trip from my home and it picks up the route with the steepest possible climbs, even if I take "Couch potato" as "fitness" choice. If I try to nudge it into another general direction, it will pick up again the hardest route in that direction. I know for having ridden them, that there is one route completely flat, another with just a few very small hills, and a few other ones with one or two longer but smooth climbs.
It seems to just aim at the shortest route, doesn't matter if there are 10% climbs in it.
For someone who has done his faire share of bicycle touring using Google Maps, I can't count the number of times it has taken me to barren fields in the countryside pushing my bike through mud. It even got me almost arrested once in France, taking me riding on the shoulder of a 110km/h highway with no exit in sight.
For anyone getting into touring, Komoot is much better for cyclists.
I actually didn't get "on" the road, I stopped before the insertion lane was over, and cautiously walked back 200 or 300 m to the entry point (because of course this part was one-way, and the last thing drivers expect is meeting a cyclist there, in a forbidden direction, especially right when they prepare to speed up to insert into highway traffic). I looked for the warning/interdiction sign I had probably missed: there was no sign. And you check any entrance on Streetview if you have time to lose, you won't find any.
Worst thing? I had to take that road anyway a bit later (farther), because it is the only bridge around, and I was late. I had already taken that horrendous bridge in the morning and I had sworn I would never take it again, but I had no choice.
See how narrow both the bike lane and the car lane are? And when I was there it wasn't practically a car lane but a lorry lane. Speed limit is 70 km/h but all of them overspeed. And anyway, the bridge is damn high, so you are (well, I was)
stuck at 15 km/h in your little painted lane with a wall lorries passing 50 cm away 50 or 60 km/h faster. Of course, over this kind of bridge, you may add some wind, possibly sideways, to fully enjoy your cohabitation. And then you go down, still in your 70 cm wide lane, as it is as steep as when you climbed, you reach 50 km/h, perhaps 60 km/h, and all of sudden, first you meet a large hole in the road where the bridge deck articulates with the ground part, and right after you jump over the hole, the bike lane turns right and the only possible way to turn is to come to basically a full stop. Then you stop for a while to calm down.
Seriously. I have been waiting for a "public transit + bike" routing option for a decade at least. You know, where public transit is an option but using a bike to transfer between stations and destinations is preferable to driving or walking. It appeared finally sometime this year, from what I noticed.
Also, pleaassee allow us to set the language of the text-to-speech manually on android without having to switch the lange of the whole OS. In the Netherlands, you are no longer allowed to hold your phone in your hand or look at it while riding a bike. Purely going by the English text-to-speech for Dutch streetnames is very frustrating, whereas translating the whole OS to Dutch just looks silly (UI concepts work much better in English).
(I guess these examples are only salient for Dutch speakers, but I'm 100% sure that similar cases exist for all other languages, where people don't want to switch the language of their whole OS just to hear streetnames spoken in a non-....stupid way)
For locals, having English pronunciation sounds absolutely ridiculous.
But for tourists, local pronunciation can be utterly unrecognizable. ("Rio Tinto" becomes "Hee-oo Cheen-too".) The mangled English pronuniciation, at least, is mangled to what a tourist can recognize/expect.
But I absolutely agree that a toggle would be a super-nice feature. As well as phonetic hints when a road has a different pronunication from what you might guess (especially in English).
Yeah, I understand the problem. I feel like all these UI design decisions are mostly due to a lack of recognition that language-use is not binary. Auto-correct used to be absolutely unusable if you mixed typing in multiple languages on your phone. Now that language toggling is easier it's only really a problem when you mix languages in the same sentence, mixed-language models do not seem to exist (yet?).
So yeah, tourists should not be forced to hear Dutch streetnames in Dutch, just as I shouldnt have to change the language settings of my whole phone to hear them in Dutch.
Several apps using OSM seem to handle the toggling just fine (e.g. Maps.Me), and apparently the cycling experience is superior on OSM anyway. I used it solely for the offline maps, but maybe I might start using it as my main maps provider.
> mixed-language models do not seem to exist (yet?)
They have since 2016, at least on iOS. [1] I thought it would be a godsend... but I ended up turning it off, because if I type 98% in language A and 2% in language B, it changed words in language A to language B far more often than made up for getting words in language B right.
Google Maps, from my experience, is unusable for cyclists - I don't know where they're getting their data, but one instance that stands up to me till this day is when it tried to get me on an express way, on a lane with opposite traffic. Or, when it tries to get me to go through pathways that are looooong overgrown. That's in London.
Agreed, in Germany half the forest paths near me are missing. It isn't rocket science, the crowdsourced OpenStreetMap data has them. (And Google would be welcome to use the data for the price of adding credits, since it's open data.)
Does Google do blue-sky software projects like this for the good of society anymore? I've heard rumors they've gotten more focused, I've seen their offerings shut down, and I am not sure this project would actually be profitable for them.
Google invests 10s of millions in their nonprofit google.org. They work on some software projects. Google Arts & Culture also comes to mind, I don't believe they make any money from it.
Do you think bike lanes are a good project for Google.org to sponsor? Is a nonprofit-sponsored collaboration within an existing product like Google Maps actually workable?
I stopped using Google maps for turn by turn directions on my bike after every unfinished road told me to make a right followed by an immediate uturn for 10 miles. Maybe they should buy Garmin. They have the best data set in the fitness industry.
Google already placed their bet by buying Fitbit. There is zero chance they will buy Garmin. But Garmin might be willing to license their cycling heat map data for enough money.
As a frequent cyclist these are generally good suggestions. Google Maps is helpful for planning bike routes, but the biggest problem I've found is that it doesn't differentiate between surface types. Some of the routes it recommends are really only usable on a mountain bike and not suitable for a road bike with 23mm tires.
I prefer to use Garmin Connect most of the time because it does differentiate between road bike versus mountain bike courses, and can do popularity based routing for either type. But sometimes the routing algorithm goes crazy and gives you routes that make no sense at all.
Strava data is available for license without acquiring the whole company.
Somewhat tangental:
There used to be a website using google maps called roadbiketoaster that allowed you to select sections of road and strung them together as a route (instead of having to select points). It would then give you an elevation profile, total distance and a cue sheet detailing each turn. I think it went away due to some google API issue and I haven't found anything that can do something similar. Does anyone know of a comparable solution and/or what google changed about its API?
If you are searching a good offline bike navigation app for Android and iOS check "Bike Citizens". Data is based on OSM. Works without internet connection. You pay per city once. Focused on Europe and on cities, but available around the world:
How's that different from looking at a digital map and planning it yourself? I personally trust navigation most of the way (I'll double check where it tries to send me roughly, but I don't look at details), but I always check out the destination and put the marker on a place where I see or expect parking to be available instead of blindly going to the address. If you add waypoints on a digital map instead of using paper, you can still have turn instructions, time planning, distance measurements, etc. yet still plan how you like to drive.
Nothing against paper maps, just curious if there's a reason why that wouldn't be best of both worlds.
Directions for cyclists are just that much worse. On a long tour, unless you spend the time looking into every intersection, you may end up for a couple miles on a highway/expressway/avenue with cars going 40-60mph, and scaring the crap out of you. This isn't an argument against digital maps of course (you can still make a track and overlay that onto a map), but just wanted to add a bit more context.
Does that also happen in well known apps like OsmAnd or any of the OpenStreetMap.org default routing engines in an area where the roads are tagged correctly as car-only roads xor cyclepaths are available? Because that sounds like a bug we can fix.
The issue with OsmAnd and OSM.org in the US is often that roads are either not tagged properly or they are tagged as cycle paths because the local government has designated them as a cycling path, but the street itself is quite unsafe. There's no real problem with the path routing libraries, a lot of it is just the data.
I've thought about exploring an alternate set of tags to tag something in OSM as a "popular" cycle path, but I'm not sure how to go about proposing that tagging system.
Google's support for cycling is very location-dependent. I used it to plan a 16 day trip from Berkeley to Crater Lake and back in 2015. It was invaluable.
From my point of view the best feature is that Google shows elevation gain, which is a huge deal for a bike. I've checked this multiple times and it seems quite accurate.
The worst feature is that some routes simply aren't shown. Google will not show a bike route from Grants Pass Oregon to the coast going directly over the Coast Range. The best route is over Bear Camp Road but Google no longer shows it. People have died on that road--most famously in in 2006 but there have been others. [1] The route it does show is suicide for a bike. (Try Grants Pass to Gold Beach to see this.)
I usually pretend I'm walking when I use Google Maps in Amsterdam. Most of the time that's the shortest route on a bike as well or close to it at least. No hills and lots of bicycle paths here luckily :)
> It doesn't appear to factor in aspects unique to cycling, for instance difference in elevation, when calculating the best route.
This quote is followed immediately by a screen capture showing the text 'Mostly flat' underneath the journey time and distance.
In fact in Edinburgh, UK my experience with Google Maps has been that it automatically selects routes that use more cycle paths, roads with cycle lanes, and flatter or downhill options. Maybe this is different per city?
I get that cycling is complex to route but when I ask Google Maps for directions to the city centre from my apartment, it sends me over a huge row of stairs.
I notified them of the issue a couple of years ago and it got fixed a few months later, but it came back again. At that point I gave up as I don't really use the service for cycling anyway. I know my city well enough and when I cycle outside of it, I use Strava.
I think most cyclists tend to ride closer to home, where they know all the routes so wouldn't need Waze Vs going on long trips in a car where you don't know the route and the road is more likely to be blocked of by an accident/traffic (when did you last see a traffic jam of bike's?)
I don't know all potential routes in a 20 mile radius of my home, no.
In a large metropolitan area, there are literally many hundreds, if not thousands of miles of roads within such a circle.
And something changes every year. Sure, someone who does nothing but ride, day in, day out, like a bike equivalent of an old school taxi driver, would learn them all after some time. I don't have that amount of time to spend.
I don't see traffic jams formed by bikes, no, but I regularly see construction crews regularly closing bike lanes and bike paths for work, for parking their construction equipment, or for setting up temporary signage.
There are a couple other mentions on this thread of OpenStreetMap, but what they don't mention is that half the value of google is in their mobile app. There are a few mobile cycling apps, mine included, that solve this while using OSM data for routing.
We use an open source routing engine called graphhopper. It's really good, the team behind it is great, and I strongly recommended it out of all the major routing systems. We tried them all, graphhopper is the most extensible, fastest, and best managed project. Graphhopper takes OSM data and computes a routeable graph based off one or more routing profiles you specify. There are default profiles, but we have spent a considerable amount of time making our own custom routing and weighting profiles to give the best directions for cyclists, and are in the process of extending it further so that it takes into account our database of about 50mm recorded bike rides by our users. The result is that our cycling directions are pretty dang good at this point. We hope to deploy the new popularity based routing stuff in the coming weeks.
We have both a web and mobile route planner that are worth checking out. The mobile route planner is a paid feature, but allows free demos without saving. The web route planner is free, though some of the more advanced bits (multi-route editing, advanced libraries of routes, etc) are paid.
Finally, another mention in this thread of a quadlock - they are awesome. I take my quadlock mountain biking and it has never budged. I'm not talking gentle mountain biking, but big jumps and drops and high speeds. They are super durable, the case is solid and is extremely protective of your phone. They are worth the premium!
One big issue is Google ignoring good free data sources. In Vienna all cycle lanes are available as ODG data. The license is very liberal, just mention the source. Instead Google Maps shows old and wrong cycle lanes.
Sigalert does. Might be US only and only knows about freeways and big roads. I use it (in Los Angeles) to check the freeways before I use Google Maps to get my route.
Yes, as long as they're not noise cancelling I don't see what the issue is. As for the bike mounted cell phone you're not staring it the whole time or interacting with it, it's the same as a GPS display in a car.
There have been experiments where cyclists with music on headphones were more likely to hear a signal than drivers behind closed windows, without music and with the engine off. Certainly depends a lot on headphone model and volume setting, but it doesn't have to be a safety issue (unless your expectations are so high that you consider closed windows an issue as well)
The sound quality is not great for music, but they are perfect for voice. The big selling point is that they don't cover your ears, so it doesn't affect your ability to hear your surroundings.
Even more surveillance capitalism, now even on my bike? No, thanks. I'd rather see more community- or municipality-driven efforts with an ethical, open approach to data and customers.
I recently started using my bike as 100% of my personal transportation. The first time I went to my studio, I used GMaps to get me there. It took me through some very unsafe parts of town even though the normal route would have been so much safer. In this case, safer in the sense of getting jumped/mugged vs traffic safety. I trusted it the first time just to see wtf it was doing. Won't use it again, and went back to just using common sense.
I suppose that depends on if they can say "we are using police data go ask the police about this, we just want to keep our users safe" as well as how much they care about such headlines .
Even in the Netherlands, it's often better than Google Maps. And it can serve as the basis for apps building on top of it like https://cycle.travel/.