A DPR is only as good as the survey underneath it, and most are written from memory.
The document is the deliverable, but it is not the work. What decides whether a DPR survives scrutiny is whether every number in it can be traced back to a household surveyed, a boarding counted or a corridor measured — and whether the same data can be re-run when a reviewer disagrees with an assumption.
From survey to costed plan, without re-keying.
Each stage hands its output to the next inside one platform, which is the difference between a DPR you can defend and a DPR you can only present.
Primary survey
Household, boarding-alighting, origin-destination and stated-preference instruments, collected offline in the field, de-duplicated and analysed where they were captured.
Secondary data
Census, employment, land use and existing network imported as analysis zones rather than transcribed into a spreadsheet.
Diagnosis
Coverage, accessibility and opportunity measured over the real network with barriers respected, so a gap is a gap and not a circle drawn round a centroid.
Design
Routes, stops, feeders and transfer nodes drawn against demand, with alignments snapped to roads a bus can actually drive.
Quantification
Fleet, crew, kilometres and cost derived from the design, annualised against the operating calendar rather than estimated.
Handover
The scenario locked with its assumption set versioned, so the figures can be reproduced months later when someone asks where they came from.
Every figure has a lineage.
Assumptions are versioned
A ridership figure is attached to the assumption set that produced it. Change an assumption and the platform re-runs rather than asking someone to remember what was different.
Scenarios are comparable
Baseline and proposal are quantified the same way, so the difference between them is a measurement and not two documents written months apart.
The method has a name
Clustering, coverage and routing are classical algorithms with names that can be stated in a technical proposal, and the same inputs give the same answer.
What a platform cannot do for you.
Bring the terms of reference.
The scope wording decides what the study actually has to produce. We would rather read it than estimate from a summary.