Draft requirements document for Design Contest for open sourced Bus Routes Management System for PMPML

[Nikhil VJ]

Hi Friends,

Kindly take a look at this and suggest if you have any ideas. I'll be pitching it to PMPML over the next few days.

The subject line is quite long, here it is again:

Draft requirements document for Design Contest for open sourced Bus Routes Management System for PMPML

PS: In case the attachment got lost on the way, you can get it online here:
https://docs.google.com/presentation/d/1FuLHS84uHuxQEsEquiWxcuiUdkv7Z8l29ImsYay5SHc/edit?usp=sharing

--
Cheers,
Nikhil
+91-966-583-1250
Pune, India
Self-designed learner at Swaraj University <http://www.swarajuniversity.org>
http://nikhilsheth.blogspot.in

[Amiruddin Nagri]

Did you have a look at Google Transit Feed Specification ? https://developers.google.com/transit/gtfs/reference?csw=1

Allowing data export in this format allows Bus routes to be searched using Google Maps. Also, this has become sort of a unspoken standard and many independent apps being built using this format. 

I haven't looked for it, but I think there should be an Open Source Bus Routes Management System that allows maintenance and export of data using this format. So, the contest should also encourage customization of one such app for PMPML.

[Amiruddin Nagri]

Finally spotted GTFS reference in one of the bullet points. So ignore the point about GTFS.

[Nikhil VJ]

;) :P Thanks Amiruddin! We're catching the bull by the horns here. People keep yelling GTFS GTFS so much, now let's get down to actually doing it.

Please also see this: https://datameet.hackpad.com/Public-Transport-GTFS-Format-Analysis-soh7vdzRsW5

--
Cheers,
Nikhil
+91-966-583-1250
Pune, India
Self-designed learner at Swaraj University <http://www.swarajuniversity.org>
http://nikhilsheth.blogspot.in

--
Datameet is a community of Data Science enthusiasts in India. Know more about us by visiting http://datameet.org
---
You received this message because you are subscribed to the Google Groups "datameet" group.
To unsubscribe from this group and stop receiving emails from it, send an email to datameet+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Nisha Thompson]

Hey 

Wow! This is really interesting.  I've added our transport committee (Suvajit, Srinivas, and Tejas) to the mail.  

Where did this come from? Is this something Pune is interested in hosting? Or you are asking them to?

I would suggest for contests to look at a good example of working with government sites - Design for America - https://sunlightfoundation.com/blog/2010/05/26/design-america-winners/

I would also say that doing the whole system as a contest is a lot of work. Would they have to design the entire web interface or just parts of it?

Nisa

--
Datameet is a community of Data Science enthusiasts in India. Know more about us by visiting http://datameet.org
---
You received this message because you are subscribed to the Google Groups "datameet" group.
To unsubscribe from this group and stop receiving emails from it, send an email to datameet+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Nisha Thompson

DataMeet.org

ni...@datameet.org

skype: nishaqt

mobile: 962-061-2245

[srinivas kodali]

Nikhil,

You have almost covered everything which is needed for a GTFS editor. I can add things on operations side like fuel consumptions and distance traveled, but it will complicate things. 

Regards,

Srinivas Kodali

[Nikhil VJ]

Hi Nisha, Srinivas,

Thanks.. the idea started out with engaging a college's whole computer engineering department or a prof and his/her students for a full-fledged internship/mou project that extends over at least 3 months. And then I thought it would be better to throw the challenge out in the open instead of locking-in dependency... if the bus authority is the one issuing the challenge then it'll probably carry through and get taken up seriously. Professors usually complain about not having meaty projects for their students' internships. So let's meet that demand.

So right now this is how I envision a possible contest timeline (I'm thinking to much ahead but here it is anyway):

1. Announcement with open sharing of data and design requirements (so really kicking it off with all resources made available)

2. Taking in registrations from teams

3. Forming an online community of all participating teams and airing out all doubts, clarifications etc there

4. Setting periodic "checkpoints" of select features that each team must achieve and showcase or else be dropped from the competition. (like achieving map display, achieving stops-editing, on-map editing, GTFS conversion, etc)

5. Final lap of 1 month to get everything together, and have to put up fully functional prototype loaded with data for public review by a set deadline.

6. Some surprises that I don't want to reveal right now, they'll work only if kept secret.

And here's how I'm thinking of bringing the aspects of collaboration in :

Teams can contact each other and merge into one, with no roll-back possibility. This will be easily done IF we haven't kept any prize money that needs to be split. This could lead to, say, a joint solution created by 2 or 3 colleges working together.

I don't care if it doesn't work out at the end.. something of this sort will produce something or the other that's useful. After all, when you want to hit a target, you need to fire a few degrees upward to account for gravity.

What I've visualized in the ppt is basically a visualization of the needs that the folks at PMPML have been telling us again and again over the last few months. 2 weeks back the newly appointed CMD also told us that they desperately need this kind of thing, and so I decided to flesh out the requirements so we have something detailed to go with.

This would happen at the output end of an ongoing effort to (finally!) standardize the data that they have, bringing it to a format that is both human and machine compatible (and PMPML-approved.. their consent and commitment is critical). So it's taking shape here, please see the "swargate" sheet:
https://drive.google.com/open?id=1ppFJeb7Dnj6-1yvniH2Q6exQFzK6fM0wsZNn4XZIvkI

Base data prepared by an org named ITDP, this data is a few years old but is more recent that the GTFS file we have, and has had ground-truthing and stops rationalization done. So the plan is to populate this, and then guide the PMPML's transport managers through editing and updating it, and simultaneously we might launch the contest.

[shirish शिरीष]

at bottom :-

Hi all,
Was talking to some people and came up with few links which could be useful :-

https://mapzen.com/projects/valhalla
https://github.com/valhalla

https://transit.land/
https://github.com/transitland

http://transit.geotrellis.com/#!/travelshed
http://transit.geotrellis.com/travelshed.html
https://github.com/geotrellis/geotrellis-transit

If anything, it tells that there are more than a few options besides GFTS.

--
Regards,
Shirish Agarwal शिरीष अग्रवाल
My quotes in this email licensed under CC 3.0
http://creativecommons.org/licenses/by-nc/3.0/
http://flossexperiences.wordpress.com
EB80 462B 08E1 A0DE A73A 2C2F 9F3D C7A4 E1C4 D2D8

[srinivas kodali]

Shirish,

Transitland is just a data warehouse for GTFS datasets and Geotrails is good for identifying transit reachability factors. But you need perfect GTFS files to use Geotrails. 

The issue is whether to adopt GTFS or not. Its more of creating these datasets is a time intensive task. Its very hard to explain this to a non programmer sometimes. 

What we need is a much more simpler solution to export to GTFS if pre recorded data exists. I would like anyone else to not occupy the thread with GTFS and use the existing hackpads.

Regards,

Srinivas Kodali

[Craig Dsouza]

Hi Nikhil,

Been going over the presentation and first of all this is great work! Very systematic thinking. 

I have added some of my thoughts below. Not all of this is necessary in v1.0 of this Bus routes mgmt system, but thinking long term, these are features that i thought were important.

Stop viewing

• Let users rate/comment on the state/quality of bus stop. 
  
• Give the users the timings (departure/arrival) for their starting bus stop to the stop they want to go to (some bus maps i've seen only show what time the bus leaves from its depot so the user does not know what time it will reach his/her own stop) Maybe you could have an option for 'Leave now' or 'Leave later' like Google maps

Route viewing

• Let users comment on the state/quality of the bus/driver/conductor. Something like the olacabs app. (only makes sense if more or less the same bus drives on the same route each day)
• Let users see a tab showing the expected and actual arrival times of the bus (on average, for different days of the week) sort of like what the irctc website has for trains. This could be enabled with an app that records bus passenger gps data. I have an irctc app that does the same when i'm on a train

Route management

• If the system could somehow show live data for 'number of passengers' on each trip (based on ticket sales/avg distance travelled) that could form the basis for more responsive/flexible decision making on which routes need more frequent buses. It would also give insight into whether the use of buses in the city is increasing/decreasing, thereby giving policymakers incentive to improve the system (because metrics show concrete improvements)

Besides the above points we could also have social media integration, so users can share any positive experiences.

Also with this system the mobile app would be as important if not more important than the web interface, most internet traffic these days is mobile.

I also had some thoughts on the discussion and proposal put forth by Adhiraj. As a separate initiative, it can very well be made the goal for a modelling exercise, seems like a problem that programmers could do well with. The goal would be optimizing the routes for 'time taken' and 'frequency' based on multiple inputs.

Regards

Craig

On Tue, Jul 7, 2015 at 5:08 PM, Sarang Mahajan <sarangk...@gmail.com> wrote:

Dear Nikhil,

A similar exercise was conducted by Mumbai Environmental Social Network for the Bus route mapping of BEST buses. It was not a competition but a project for MMRDA. It included pasting of maps on the bus stops which contain related information of all the routes passing through that stop. It has been a successful and recognised project by MESN. I am attaching a photo for your reference. It also won the VOLVO sustainable mobility award!

Please let me know if any such method can be incorporated.

Regards,

Sarang K. Mahajan

Mobile: +91 9011954500 

--
You received this message because you are subscribed to the Google Groups "PTTF" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pttfgen+u...@googlegroups.com.
To post to this group, send email to ptt...@googlegroups.com.
Visit this group at http://groups.google.com/group/pttfgen.

[srinivas kodali]

Craig,

All the points you mention will come into applications which are commuter focused, which can developed too but as a different application. The current issue is to develop a system fro depot managers and operations people.

Regards,

Srinivas Kodali

[Craig Dsouza]

Srinivas,

Ok no issues, I guess I misunderstood the target audience. So I take back the first two points, but perhaps the 'Route Management' point is relevant to the depot managers too.

Regards

Craig

[pallavi palsokar]

Hello everyone,

This is great work Nikhil! We commuters really need something like this.... Badly... :( 

There are a few questions that come to my mind... - Are we looking at a system that would work on real time data or stand alone? Will the system get the real time updates of the buses? Where they are? At what time they reached a spot etc? It the system being designed considering this? or it is a tool which would help PMPML to plan their routes and have inventory of bus stops? I would like to be clear on what exactly we mean when we talk about 'Management', to what extent are we talking about managing the system? Once I am clear on these, will have more thought to share :)

About the presentation....I see the presentation here has covered all the important aspects that should be looked at but I feel that its has not talked much about the 'Analysis and Statistics' part of it. I see a 'Analysis, Statistics' Tab present but there is nothing written about it. Can we/ should we give few ideas on what 'Analysis' should/ may include? If one is developing a 'Bus Route Management System' - Managing would require a set of data and the system would collate the data and good analysis support would 'improve' the functioning of the system. Again here I am not able to understand the actual functioning of the system... once I am clear on that would love to share few ideas on analysis too...

Another thought that comes to my mind after reading Chirag's mail which was written considering more about the commuters perspective, we may also have some kind of commuter feedback to the management system that may help the system managers to improve the service delivery. Effective management of the system is effected by many other aspects other that route and stop management.... issues like: buses not stopping at designated stops, over occupied buses, bus breakdown frequency, driver/ TC behavior etc are also important. With increased used of smart phones creating this link should not be difficult....

Please share your thoughts on the above comments/ suggestions.

Regards,

Pallavi