Riding an ebike has dramatically changed the routes I take when commuting because I now prioritize safety and enjoyment over pure efficiency or avoiding inclines. My commute is 25 minutes on an ebike vs 70 on a regular bike, or 30 in a car in traffic.
Also something to consider is that I can happily cycle 2-3 hours on a flatish route on a normal bike, but will find hilly routes something I do not do.
On my eBike I'll happily explore hills without thinking about it. Given that my eBike is pedal-assist, you still have to work hard to get a top speed going up a hill but I no longer feel hills are a problem.
Barely related but EV drivers also change their path. I don't have an EV (car or bike) don't really like driving in traffic (too chaotic for my brain) I often take smaller, slower roads where I just coast along smoothly. That said I wish I could try an ebike or even an EV.
It's an interesting problem because there are so many factors and even the goal isn't constant from day to day.
Goals
- get from A to B as quickly as possible
- get from A to B as safely as possible
- get from A to B as enjoyably as possible
Factors
- fitness of rider
- type of bike (Riding a fixed gear bike up a steep hill is bad for enjoyment, but riding a mountain bike up a steep hill is not)
- weather (headwinds, tailwinds affect time, affects safety of certain routes as well - don't tell me to ride down a steep windy hill when there is a possibility of ice)
- time of day (affects safety due to light levels)
- traffic volume
- number of lights/stop signs (this affects fast riders more than slow ones because of the effort wasted accelerating and braking is greater)
- number of turns (turns make the rider slow down and wait for traffic and also increase the complexity of following a route)
- scenery/environmental factors (eg. ride through an industrial area or through a waterfront park)
And it's also not said that all these factors have a linear impact on the Figure of Merit. I guess that this might be a nice application for a machine learning classifier, to rate a route's ride quality depending on all of these inputs (and more.)
NB: "Hybrid" bike in this context (probably) means a city/commuter bike, not an electric/human hybrid. "Hybrid" bikes used to describe upright bicycles with flat handlebars like a mountain bike, but less aggressive and narrower tires more suited to commuting than to trails. These days the industry is calling them "fitness bikes"; one presumes to avoid confusing people who think of "hybrid" as meaning battery-powered, eg in the automotive context.
Wonderful. I'd really like these same features in Waze and Google Maps - if it means I can avoid having to cross a busy intersection that's not a 4-way stop, and reduce stress and increase safety, I'd gladly accept a route that's a few minutes slower.
Yeah, this is a serious problem with Google Maps directions. They want to send cyclists down the fastest routes, which are usually the least safe routes. There's a route in Kent, WA that Google was recommending that was downhill on a 45 MPH two-lane street with no shoulder or dividers. Meanwhile, about a mile further south there was this safe, car-free route that wound through residential subdivisions.
It's not GMaps, but I've been using Bike Citizens and it has an option to select one of three profiles, depending on how much you want to avoid busy/larger streets (Note: as far as I know they don't use Mapzen's algorithm.)
Only a tangent, but I just started a new job and used Google Maps to find me a route and it's absolutely glorious. It's on roads I never would have thought of, avoids streets with heavy traffic, only has 3 traffic lights.. I'm not sure if it's just dumb luck, or the fact that there are a decent number of neighborhoods through Palo Alto that are relaxed tree-lined streets, or what.. but it's only a couple miles shorter than my commute to my old job and an order of magnitude more pleasant. Looking forward to using this API to see if it recommends changes.
I happen to know those intersections. They could have improved that wiggle route simply by rating the intersections. I'm not sure but it looks like they were not doing that.
The old route crossed two stop signs and did a left turn at a traffic signal. Left turns at traffic signals really suck.
The new, proper wiggle route is two left turns and a right turn, all at stop signs. Left turns at stop signs are not that much worse than going straight across -- and vastly better than a traffic light. Right turns are practically free.
So by simply rating the difficulty of intersections using a global metric, it seems like they could have made the same wiggle improvement without getting into rating each section of street. My point here is not that assessing things like accomodation_factor and road_stress are useless. Rather it appears that they've overlooked the much simpler, global improvement of rating intersections.
Only tangentially related to your post, but, two new traffic lights are being installed on the Wiggle. Haight is getting nine new lights, two of which are at Scott and Pierce. Maybe people will wiggle slightly differently to avoid the left turns at the lights.
Is there any way to see a demo of this algorithm for a route of the user's choosing? I'm not familiar enough with San Francisco to evaluate the improvement mentioned in the blog. If I could try some routes in Atlanta, however, I'd get a better sense.
The screenshots in the article would be more helpful if they were zoomed a little closer, at least enough to show all the street names being mentioned. A topo or shaded relief map would also help.
I'm glad people are working on this. The bike directions on Google maps aren't very good, IMO
So I decided to see how to get to that nice Pho place across town .. 600 Washington Ave
Philadelphia, PA 19147 or the one next door at 610 Washington Ave
Philadelphia, PA 19147 .. in both cases it truncates off the street number and drops me at a wrong location. Does that mean OpenStreetMap needs some assistance in that area? Even stranger if I put one of those addresses in both source and destination it does a strange 5 block loop to get back to the same wrong start point.
A friend has been working on a similar project for a while that allowed the rider to tailor, in great detail, the kind of route required. The algorithms involved are pretty interesting!
Any chance it might eventually allow users to specify a single start/end point and a desired distance? That'd be much more useful for people who ride for sport, and even though there seem to be a few websites that generate such looped routes, they don't appear to do it particularly well.
I ride in London, and love https://www.cyclestreets.net/. There are versions for the web, android and now iPhone. It has a 'quietest route' option which I love. Android version gives you voice navigation. I put the phone in my shirt pocket and have it talk me through quiet streets avoiding the hassle of traffic. Open source. Runs on Openstreetmap data. Needs funding and promotion.
Yes -- at the bottom of the post we show how you can use our mapzen.js library to draw routes on a map. If you know Leaflet it's relatively straightforward.
Is there a bike routing product that uses this new api that I can use?
I spend hours finding the "less stressfull" bike routes.
I like hills but I really value safety.
Strava[1] provides a cycle route builder that uses the popularity of routes based on their dataset of uploaded rides. As it's using recorded rides, typically by local riders, I've found it produces good routing; quiet roads similar to those you'd choose if you knew the area well.
It's a rabbit hole, but there is pretty good support for loading OpenStreetMap data onto Garmin devices (other device makers did a better job obscuring their formats):
I would buy this if it was significantly cheaper than a phone and made it easy to adjust those tuning parameters. I prefer to keep my phone protected on the bike in case of a collision / fall.
Leaflet is for requesting and displaying raster tiles on the client. PostGIS is for storing and managing geo data. Leaflet and PostGIS do not talk to each other. Inbetween you need a tile renderer (like Mapnik).
Leaflet is a great library if you don't need much (use OpenLayers then).
Likely not as many (most?) sidewalks aren't officially open to bikes (Regardless of enforcement on the one hand and regardless of some people preferring to use them on the other).
Hi! I'm with Mapzen. Setting `use_roads` to zero means a preference for using roads tagged in OpenStreetMap that have bike infrastructure, as opposed to roads without bike lanes.
We don't route on sidewalks, and will avoid paths that are tagged `bicycle=no`. You can get a sense of these on our bike map. https://mapzen.com/bikes/
[+] [-] boxcardavin|8 years ago|reply
[+] [-] awjr|8 years ago|reply
On my eBike I'll happily explore hills without thinking about it. Given that my eBike is pedal-assist, you still have to work hard to get a top speed going up a hill but I no longer feel hills are a problem.
[+] [-] agumonkey|8 years ago|reply
[+] [-] OstrichGlue|8 years ago|reply
[+] [-] erentz|8 years ago|reply
[+] [-] dfrey|8 years ago|reply
Goals
- get from A to B as quickly as possible
- get from A to B as safely as possible
- get from A to B as enjoyably as possible
Factors
- fitness of rider
- type of bike (Riding a fixed gear bike up a steep hill is bad for enjoyment, but riding a mountain bike up a steep hill is not)
- weather (headwinds, tailwinds affect time, affects safety of certain routes as well - don't tell me to ride down a steep windy hill when there is a possibility of ice)
- time of day (affects safety due to light levels)
- traffic volume
- number of lights/stop signs (this affects fast riders more than slow ones because of the effort wasted accelerating and braking is greater)
- number of turns (turns make the rider slow down and wait for traffic and also increase the complexity of following a route)
- scenery/environmental factors (eg. ride through an industrial area or through a waterfront park)
[+] [-] konschubert|8 years ago|reply
[+] [-] megamindbrian|8 years ago|reply
[+] [-] rconti|8 years ago|reply
[+] [-] icebraining|8 years ago|reply
[+] [-] arikr|8 years ago|reply
Anyone from Waze or Google Maps team here?
[+] [-] jschwartzi|8 years ago|reply
[+] [-] icebraining|8 years ago|reply
[+] [-] rconti|8 years ago|reply
[+] [-] arikr|8 years ago|reply
[+] [-] abalone|8 years ago|reply
The old route crossed two stop signs and did a left turn at a traffic signal. Left turns at traffic signals really suck.
The new, proper wiggle route is two left turns and a right turn, all at stop signs. Left turns at stop signs are not that much worse than going straight across -- and vastly better than a traffic light. Right turns are practically free.
So by simply rating the difficulty of intersections using a global metric, it seems like they could have made the same wiggle improvement without getting into rating each section of street. My point here is not that assessing things like accomodation_factor and road_stress are useless. Rather it appears that they've overlooked the much simpler, global improvement of rating intersections.
[+] [-] zten|8 years ago|reply
[+] [-] sathomasga|8 years ago|reply
[+] [-] maxerickson|8 years ago|reply
http://www.openstreetmap.org/directions?engine=mapzen_bicycl...
I checked the request, the JS powering the OSM feature isn't sending the parameters so they are probably the 0.25 defaults there.
[+] [-] jlarocco|8 years ago|reply
The screenshots in the article would be more helpful if they were zoomed a little closer, at least enough to show all the street names being mentioned. A topo or shaded relief map would also help.
I'm glad people are working on this. The bike directions on Google maps aren't very good, IMO
[+] [-] schemathings|8 years ago|reply
[+] [-] maxerickson|8 years ago|reply
http://www.openstreetmap.org/#map=19/39.93668/-75.16656
[+] [-] dve|8 years ago|reply
http://zikes.website
[+] [-] phumbe|8 years ago|reply
Any chance it might eventually allow users to specify a single start/end point and a desired distance? That'd be much more useful for people who ride for sport, and even though there seem to be a few websites that generate such looped routes, they don't appear to do it particularly well.
[+] [-] fcc3|8 years ago|reply
[+] [-] lasryaric|8 years ago|reply
- traffic - elevation gain - type of street - a way to represent the distance overhead compare the other segments.
Any framework / dev env where I could easily get that entry point and see the top results on a map?
[+] [-] burritojustice|8 years ago|reply
[+] [-] lasryaric|8 years ago|reply
[+] [-] JofArnold|8 years ago|reply
[+] [-] slashdotdash|8 years ago|reply
[1] https://www.strava.com
[+] [-] loeg|8 years ago|reply
[+] [-] thisiscool|8 years ago|reply
[+] [-] maxerickson|8 years ago|reply
http://wiki.openstreetmap.org/wiki/OSM_Map_On_Garmin
I guess the routing enabled bike maps will take advantage of whatever data is present.
BBBike supports custom areas:
http://extract.bbbike.org/extract.html
[+] [-] blt|8 years ago|reply
[+] [-] brightball|8 years ago|reply
[+] [-] aw3c2|8 years ago|reply
Leaflet is a great library if you don't need much (use OpenLayers then).
[+] [-] novia|8 years ago|reply
[+] [-] Facemelters|8 years ago|reply
[+] [-] maxerickson|8 years ago|reply
[+] [-] burritojustice|8 years ago|reply
We don't route on sidewalks, and will avoid paths that are tagged `bicycle=no`. You can get a sense of these on our bike map. https://mapzen.com/bikes/
[+] [-] vermooten|8 years ago|reply
[+] [-] drumttocs8|8 years ago|reply
This kid is pretty cool.