CityCyclingEdinburgh Forum » Infrastructure

Portobello Promenade

(105 posts)

1234

  1. Edinburgh Cycle Training
    Member

    The biggest issue IMHO is the small but very vocal anti cycling lobby who inhabit the Porty People FB page in particular. I honestly would not be surprised if they formed a vigilante group and attacked cyclists or damaged parked bikes!

    I've been a 'local' for over 20 years and ride or walk the prom almost daily. I often see inconsiderate riding but can't recall one single instance that I'd consider 'dangerous'

    By and large, folk mix well and get on. For every inconsiderate cyclist I see however, I see more inconsiderate pet owners, and even more parents who let their children wander aimlessly on a mixed use path.

    A few signs reminding cyclists that the prom gets exceptionally busy on 'nice' days so an alternative route might be better and that on quieter days dog and child walkers should be aware cyclists might be going quicker than they expect, would not go amiss.

    Posted 9 years ago #

  2. Murun Buchstansangur
    Member

    I'm afraid I've given up cycling along the prom, which is a shame, as it is quite lovely most of the time. Unfortunately its design (walled blind corners to access thoroughfares) and unpredictable bi- & quadrupedal usership began to render the whole thing more of a hazard perception test than a pleasure, at least to me. Down the High Street & onwards for me now. Just means the prom businesses & indeed Porty altogether misses out on my coffee/cake/beer patronage.

    Posted 9 years ago #

  3. tk
    Member

    I hadn't read that far down in Maureen's latest email but will certainly drop her an email.

    The vocal anti cycling lobby made some pretty nasty personal remarks to individuals on the recent threads (which were deleted by the admins) and seem to be taking the discussion to a closed group called 'Porty Issues'.

    This is the same people who complained about cyclists using their bells and said they would never move out the way for a cyclist (slow or fast)

    Posted 9 years ago #

  4. Dave
    Member

    "I would say that the absolute maximum acceptable speed on NMW is 15mph at any time of day or night.

    Just because you think it's clear/deserted doesn't mean that there isn't a drunk, a dog or a child lurking in the trees that may run out, or something else unexpected."

    This sounds reasonable, but when I go for a run (on foot) I can do over 10mph, so you're saying the *absolute maximum* speed on our just-upgraded flagship national cycle infrastructure is roughly walking speed faster than a runner? Doesn't that seem a bit... puritan?

    In fact it's worse than that, because a genuinely fast runner can do pretty much 15mph (after all, the best marathon runners go at 13mph for two hours), so what you're really saying is that our flagship national cycle infrastructure, purpose-built, is only suitable for going *slower* than people can travel on foot?

    "It reminds me of a very good road safety advert of a motorist going round a corner too fast and meeting something unexpected. 3 times he/she manages to avoid it. On the 4th occasion it ends in disaster."

    Isn't the key point that the motorist is going round a corner? Regardless, isn't it a bit fantastical to say that a two second gap in traffic is safe (when your car weighs 2+ tons), but a much bigger safety margin on a bike is unsafe (when your bike weighs ~8kg)?

    I can understand the argument that drivers should be more careful than cyclists because of the huge dangers involved, but IMO you're flogging a very dead horse to argue that cyclists should be more cautious than even responsible drivers need to be.

    Maybe a man with a red flag could precede all cyclists?

    Posted 9 years ago #

  5. cb
    Member

    I noticed on NMW yesterday approaching MMW, spray painted on the pedestrian side, "5MPH maximum", and a little further along, "Flash hazard lights".

    Presumably instructions for which ever council vehicle has decided to go for a wander through the meadows.

    Although the last vehicle I saw at that intersection was a fire engine. I wonder wonder what speed it did given it was there to put out a fire.

    Posted 9 years ago #

  6. Cyclingmollie
    Member

    I think the prom is as much a cycling destination as a through route*. The businesses there can expect increased numbers of cyclists because there is no room for more parking. What interested parties should be doing is campaigning for an extension of the NCN into the town from the station.
    *We visited the prom a lot with the kids' trailer for the beach, the amusements and the ice-cream van. I currently cycle from Mussy to Porty to shop about twice a week.

    Posted 9 years ago #

  7. tammytroot
    Member

    Yes, the cafes on the prom do very well out of cyclists!

    Posted 9 years ago #

  8. tk
    Member

    Some preliminary analysis of speed on the 500m West end of the Prom pretty much from the Kilns to swimming baths

    https://www.evernote.com/shard/s1/sh/1c79498e-9158-4b08-b739-fe995abee4f9/24252e1ad59bcdcf11f3f20d4b28bfb7

    Posted 9 years ago #

  9. tammytroot
    Member

    Good work tk.
    I take it you will pass this on to Wangi and or Maureen?
    This should make it difficult for the wingers to present cycling on the prom as dangerous.

    Posted 9 years ago #

  10. Dave
    Member

    tk, nice work. How did you extract the raw data? (I can see how you would scrape each individual's fastest time, but not access all attempts?)

    A nice follow-up would be to look only at rides that took place at a busy time (Saturday mid afternoon, or whatever).

    Traffic has a big impact on speeds, as I observed this weekend where roads I would do 40-60mph on were reduced to near walking speed by Tour de France spectating pedestrians.

    Posted 9 years ago #

  11. tk
    Member

    Dave - don't know how good you are at python but the steps went something like this

    - Apply for a Strava api key
    - Cut down the fetch.py script used in the work Oliver Nash did to only fetch the all_efforts json data from the Strava API. Oliver's work is at http://ocfnash.wordpress.com/2014/05/25/mining-the-strava-data/
    - Convert this file back to JSON using my own script.
    - Use the JSON to CSV tools from https://github.com/evidens/json2csv/ to output csv, defining a custom map file to unnest the JSON tree

    Oliver's scripts do some interesting data processing and I was hoping to use his original fetch (with some modifications to grab all the streams such as heartrate etc) then do a heatmap showing a histogram of speed vs distance where the intensity on the speed/distance axis shows the proportion of efforts within a bin

    I found a few issues with the individual efforts in doing this

    1 - The distance doesn't necessarily match the segment length. You actually see this in the all_efforts as both distances are shown. This can be when someone wobbles along the segment path (but stays within maximum bounds) or just the 1 second recording as the first and last point will not match perfectly the segment edge
    2 - Distances are referenced to the whole activity (although you can subtract the first value to easily correct this)
    3 - Due to point 1, distance is not distance along segment but distance travelled. To find distance along segment would need to use LAT/LONG and some serious maths to map to the nearest point (perpendicularly) on a curved path
    4 - Faster cyclists may not have a point in each bin to count thus skewing the data. Interpolation could be used but would need to be intelligent to avoid smoothing peak speed as the recorded data is not maxima

    I'm tempted to write this up and rewrite into a single piece of code that will generate graphs. If you only get the all_efforts it doesn't use many API calls and takes about 10 seconds. Grabbing all the effort data as a free API user took about half a day

    Posted 9 years ago #

  12. gembo
    Member

    Lovely graphic, nice data. Qualitatively alas a whining granny with a wee shopping trolley trying to make it to pound land on porty high street will feel terrorised by a 12mph Lycra lout careering towards her. (Hypothetical granny, not checked if there is a pound shop in porty)

    This is the reality of shard space as we have discussed before, unless you get off your bike push it round the other users and apologise profusely for existing you will be accused along the lines of Do you think you own the path or indeed You don't own the path ya bass.

    Posted 9 years ago #

  13. tk
    Member

    In terms of times of day

    If I define busy as 7am to 7pm, the data doesnt change much but then about 89% of rides are within these times

    If busy is 8am to 6pm, we see the cumulative line shift 1mph left. 77% of rides are within these times

    If busy is only May to Sep AND [[7-9am OR 3-7pm weekdays] or [7am to 7pm weekends]] then actually we see an increase in speed by 1mph. 24% of rides are in these times. This may be that Summer riding is faster which it generally is and looking at the histogram there are a few more peaks towards the fast end. This could be explained by a few cyclists doing Summer training and finding the prom is clear during these times - it only accounts for a few hundred Strava tracked rides since 2008 so its far from a regular occurrence.

    Posted 9 years ago #

  14. wingpig
    Member

    I applied to join the FB group in question but the admins haven't responded.
    A shame there's not any sort of codger GPS trace data source. Wonder if the demographics involved in interactions between cyclists and slower-moving-path-users have altered much round the bend by the kilns since the indoor bowling rink was converted to soft-play and gymnastics?

    Posted 9 years ago #

  15. tk
    Member

    For those interested, this is one of the effort summaries strava spits out when you ask for all the summaries on the segment. It shows the kind of data available. I'm interested in a split based on heart rate too to see speeds of those who really pushed themselves vs average

    {
    "kom_rank" : null,
    "distance" : 456.4,
    "average_watts" : 46.9,
    "name" : "Prom speed research - please do not flag",
    "start_date_local" : "2011-02-25T17:26:54Z",
    "pr_rank" : null,
    "athlete" : {
    "id" : 811473
    },
    "moving_time" : 147,
    "max_heartrate" : 166.0,
    "end_index" : 282,
    "elapsed_time" : 172,
    "average_heartrate" : 160.9,
    "id" : 3740166468,
    "activity" : {
    "id" : 14419877
    },
    "resource_state" : 2,
    "segment" : {
    "distance" : 536.3,
    "starred" : false,
    "end_longitude" : -3.107255,
    "name" : "Prom speed research - please do not flag",
    "city" : "Edinburgh",
    "country" : "United Kingdom",
    "id" : 7565976,
    "elevation_high" : 10.2,
    "private" : false,
    "start_latlng" : [55.956387, -3.114075],
    "start_longitude" : -3.114075,
    "state" : null,
    "elevation_low" : 5.0,
    "end_latlng" : [55.953495, -3.107255],
    "end_latitude" : 55.953495,
    "maximum_grade" : 8.9,
    "average_grade" : -0.3,
    "climb_category" : 0,
    "resource_state" : 2,
    "start_latitude" : 55.956387,
    "activity_type" : "Ride"
    },
    "start_date" : "2011-02-25T17:26:54Z",
    "start_index" : 259
    },

    Posted 9 years ago #

  16. chdot
    Admin

    @tk

    So of the people who like to think they can go fast (ie join Strava - though some join to record journeys rather than speed) most don't go 'too' fast on the Prom.

    Obviously there is no way of knowing how clear the Prom is when the higher speeds are recorded.

    I'm sure the average speeds will be higher on MMW (but not directly comparable because it's segregated and less 'cross traffic').

    I wonder how various bits of the canal are for speed.

    Posted 9 years ago #

  17. crowriver
    Member

    13mph seems a bit fast for the Prom, IMO. If it's deserted then fair enough, but even then there are blind junctions where a dog/pensioner/kid on scooter could appear "from nowhere".

    The only bit of the Prom where I go at any speed is the wide western stretch, beyond the car park, where the Lothian buses depot, car showroom and dog home are. Even then you need to watch out for dogs off the lead, joggers, etc. but can usually give them a wide berth due to the expansive space.

    Posted 9 years ago #

  18. wangi
    Member

    Thanks for pulling this together tk - very interesting.

    Posted 9 years ago #

  19. Dave
    Member

    13mph seems a bit fast for the Prom, IMO. If it's deserted then fair enough, but even then there are blind junctions where a dog/pensioner/kid on scooter could appear "from nowhere".

    Someone should go and tell the runners - according to Strava they're doing a hair under that speed along the prom.

    Posted 9 years ago #

  20. twq
    Member

    @crowriver - I had a bit of a tailwind last night, and lots of energy, got a new PB on the segment. It was great fun, and very quiet (3 walkers, no dogs). 27mph!
    Once I got to the Prom proper I slowed right down, it was a bit busier so I was going pretty much walking pace.

    Posted 9 years ago #

  21. fimm
    Member

    Maybe tk could mine the running data for the prom, just for comparison?

    Posted 9 years ago #

  22. tk
    Member

    If someone wants to create a running segment to match the cycling one at http://www.strava.com/segments/7565976 I'd be happy to pull the data. Likewise if there are a few others people want then give me the segment ids and I'll post csv files of the data you can then play with in excel

    Posted 9 years ago #

  23. crowriver
    Member

    The segment tk pulled up is the busiest, narrowest section of the Prom: the place with the cafe, pub, fish and chip shop, children's playground, nursery, amusements, ice cream and crepe stands, lots of benches, etc. Also the busier side streets terminate here, including one which leads to Towerbank primary school.

    So I think 13mph+ is too fast for this section, unless it's deserted. Which it may have been when these cyclists did that speed, we just don't know.

    Posted 9 years ago #

  24. wangi
    Member

    Well the NW end of the Prom, behind the garages, would be a good comparison to the busy section; see http://www.strava.com/segments/1473561

    Posted 9 years ago #

  25. Dave
    Member

    +1

    It would be interesting to see a comparison graph for that segment and the prom 'proper' at a period we'd assume is busy, like say June-September weekend lunchtimes only?

    Posted 9 years ago #

  26. neddie
    Member

    Good work tk. With scraping skills like that you could easily get a job at Skyscanner ;)

    I'd be interested to see the same graph for NMW to see how a segregated route compares.

    About the prom, I'd imagine that most cyclists have to slow down to 5 - 10mph to pass the Beach House/Bath Street especially when busy. It means the speed across the rest of the segment is most likely higher than the average for the segment.

    Posted 9 years ago #

  27. tk
    Member

    For the python users out there....


    import requests, time, sys, os, json

    access_token = 'INSERT API KEY HERE'

    extra_headers = {'Authorization' : 'Bearer %s' % access_token}

    api_base_url = 'https://www.strava.com/api/v3/'
    api_segment_url = api_base_url + 'segments/%d'
    api_segment_all_efforts_url = api_segment_url + '/all_efforts'

    per_page = 200 # Strava max

    #segment_id = 7565976 # "Prom speed research - please do not flag"
    segment_id = 1473561 # "Behind the garages sprint"
    all_efforts_fname = 'all_efforts.%d' % segment_id

    if os.path.isfile('%s.json' % all_efforts_fname) == False:
    all_efforts = []
    r = requests.get(api_segment_url % segment_id, headers=extra_headers)
    n_efforts = r.json()['effort_count']
    sys.stdout.write('Fetching all %d effort summaries for segment %d\n' % (n_efforts, segment_id))
    for i in range(1, 2 + n_efforts / per_page):
    sys.stdout.write('Making summary request %d\n' % i)
    r = requests.get(api_segment_all_efforts_url % segment_id, headers=extra_headers, params={'per_page' : per_page, 'page' : i})
    if r.status_code != 200:
    sys.stderr.write('Error, received code %d for summary request %d\n' % (r.status_code, i))
    else:
    all_efforts.extend(r.json())
    time.sleep(2) # Make sure do not hit Strava rate limiting

    json.dump(all_efforts, open('%s.json' % all_efforts_fname, 'w'))

    And the mapping file


    {
    "map": [
    ["start_date_local", "start_date_local"],
    ["id", "segment.id"],
    ["athlete", "athlete.id"],
    ["pr_rank", "pr_rank"],
    ["distance", "distance"],
    ["moving_time", "moving_time"],
    ["elapsed_time ", "elapsed_time"]
    ]
    }

    Posted 9 years ago #

  28. tk
    Member

    All I had time for over lunch was the behind the garages segment versus the prom. It clearly shows cyclists do take more care on the prom as they slow down by 5mph - 25% of their speed. http://www.evernote.com/shard/s1/sh/d383fdd5-9089-48b8-accc-384b2f57c80c/c438303e0a4dcaf17c8af667e44d4e6e

    Posted 9 years ago #

  29. wangi
    Member

    Thanks again for this tk.

    Posted 9 years ago #

  30. tk
    Member

    Runners vs cyclists - http://www.evernote.com/shard/s1/sh/ec5a61bd-d184-4b20-b875-534781f68c68/0968c8aedf466e9472f573400831f5de

    The run segment is from the old amusements to bath street so slightly offset but its very similar in profile

    Posted 9 years ago #

RSS feed for this topic

1234

Reply »

You must log in to post.


Video embedded using Easy Video Embed plugin