Project Plan


I'll approach the Tube Challenge like a consulting project, partly because I think it will be a smart thing to do, but mostly because I'm incapable of thinking in any other way. See this presentation for an example. It did not go down well.

From what I can gather from the web, the best tube challengers do not have a set route to follow:

"The best way of doing it, is actually NOT to have a fixed/set route. i.e. you get Camden Town and a Mill HIll East train is NOT there - do the Edgware Branch first. But if it IS there (or you know from the WTT's that your High Barnet train will connect), then do that branch first.
Many people get stuck in the trap of having a fixed route and sticking to it ... Fail! Have a route that gives you options, and be prepared to change it on the day, on the fly.
"


I'll be following that advice just as soon as I can figure out WTF a WTT is.

The project objective then is to complete the entire network in as fast a time as possible, and to that end the necessary deliverable will be a primary route description, complete with optional variations at certain points.

The deliverable will thus be a "playbook" covering these items at relevant stations:
  • Option list for what to do next
  • Decision criteria to be used to decide best option
  • Where to stand on platform / how to change platforms / how to exit station / etc

The project will be structured into 7 segments:

Primary Data Gathering

1. Network MapGenerate a geographic representation of the network, i.e., not the topographical map
2. Network Simplification
Combine stations where possible to reduce number of network nodes
3. Travel Times
Get travel time between nodes connected by track, and typical frequency of service
4. Additional Routes
Identify nodes that can be traversed outside of the network, e.g., by running
5. Interchange Durations
Estimate line changing times at each station, e.g., between 3 lines at Oxford Circus
6. Network Constraints
Generate a list of constraints to be modelled e.g., unusual station opening hours

Computer Programming

7. Algorithm Programming
Code a travelling salesman algorithm into VBA to automate generation of routing ideas

Iterative Solution Seeking

8. Solution Generation
Run program and identify likely routes
9. Sanity Check
Explicitly check promising routes with TFL journey planner
10. Improve Inputs
Enhance relevant input assumptions with additional / more accurate possibilities
11. Algorithm Refinement
Refine algorithm to avoid generating 'impossible' routes

I imagine cycling through 8 to 11 a number of times

Fine Tuning

12. Platform Optimisation
Identify the ideal location to board trains in order to later exit most effectively
13. Overland Runs
Familiarise myself with bus routes, bus stop locations, landmarks to navigate by, etc

Trial Runs

14. Trial Runs
Practise route variations, ensuring feasibility and refining the playbook

Decision Making Resources

15. Decision Points
Firm up the list of places where decisions can be made now that seen entire network
16. Support Structure
Enlist help of others for support on the day and help with decision making

The Challenge

17. Do It!
Rise at silly o'clock, take Imodium pills, and start recording train numbers. Get a life