Asset matching and length coverage

Calculations and supporting evidence

Cost calculations and peer comparisons · Asset matching, dimensions and length coverage

How much allocation per lane-meter?

One lane-meter is one meter of one lane. A two-lane, 100-meter funded segment has 200 lane-meters. Only explicit title lane counts supply the title-stated measure below. Existing inventory lanes supply a separate estimate; they do not verify the lanes funded by the allocation. Widening needs the additional lanes actually funded.

Dimensions recovered from matched assets

Existing asset dimensions can be recovered even when a title does not state the funded length. Road figures describe the recorded contiguous extent of the matched inventory section, which may not cover the whole section; bridge figures describe the existing bridge. They are separate from funded-project length coverage and do not fill missing project lengths or enter per-meter price flags. Counts are budget line items, so an asset may occur more than once.

Lane matching and exclusions

• Title lane-meter rate = whole allocation / (usable stated meters × explicit title lane count). It is not verified construction cost.
• Inventory proxy = whole allocation / (stated meters × inventory lanes). Road lane counts are measure-weighted when counts vary along fully covered ranges.
• Existing-asset dimensions are recovered separately from funded project length: bridge inventory span/width/lanes, or the union of road section intervals with measure-weighted lanes/widths. Gaps and conflicting or unknown measurements are withheld. Section size never fills a missing funded segment.
• Only exact printed section/bridge IDs or a complete exact road name with compatible region/DEO and unique section qualify; no fuzzy names.
• Kilometer post brackets map global K station values to section-local measure by interpolation, never extrapolation. Ambiguous or conflicting references are withheld.
• Widening needs funded additional lanes; approaches, slope/drainage and mixed scopes are withheld. Bridge span mismatch above 10% prevents the proxy.
• Separate existing-bridge asset estimate = allocation / (inventory bridge span × inventory lanes), only for rehabilitation/major repair or retrofitting/strengthening. This is allocation intensity over the existing asset, not a funded-length rate; it never fills stated length or enters high-rate flags. Replacement, new construction and widening are excluded.
• Inventory copyright is preserved. Old inventory does not establish the lane count at the FY2027 funded segment.
• Inventory proxies remain separate from title-stated rates, historical comparisons and per-meter high-rate flags. No missing funded lane quantity is invented.

Surface, condition and prior surfacing

Condition flag: rehabilitation/reconstruction/major repair only, full matched road range assessed Good/Fair with a known pre-budget assessment year. Recent-surfacing flag: rehabilitation, preventive maintenance or other explicitly described pavement work where the full matched range has a positive recorded last-surfacing year within five budget years; unknown/zero values are not filled. Widening, approaches, slope/drainage and off-carriageway scopes excluded. Bridge condition is displayed, but an unknown assessment year prevents the dated flag.

Review leads, not proof of waste or incorrect prioritization. Assessment and budget dates differ; later deterioration, structural/safety work or different funded components may justify rehabilitation. Last-surfacing inventory is older and cannot establish that no later work occurred.

Surface and condition filters apply to the cost tables below. Price peers use the same recorded surface, condition and assessment-year group when available; records without a supported match are compared only within the explicitly unknown-context group. A condition flag and a high-rate flag answer different questions. Last surfacing comes from the older lane inventory, whose latest reported years are not a complete record of subsequent works. The five-year threshold is a disclosed screening choice, not an engineering service-life standard. Bridge construction, edit and copyright dates never substitute for a condition assessment date.

Lane tables and charts below have program filters, sorting and CSV exports. Estimates do not enter the existing per-meter high-rate flags or historical comparison. Named overpasses, flyovers and viaducts now use bridge peers, including Burol Overpass; ordinary road peers are not appropriate for those structures.

Inventory sources and vintage

The matched inventory ID column shows the supported road section ID or bridge ID, rather than a budget occurrence ID or an ArcGIS object ID. A dash means no supported inventory match. Asset identity does not verify funded quantities or district attribution.

Download all lane matches, rates and exclusions (CSV) · Download full lane evidence, methodology and source hashes (JSON)

130 more line items now have usable stated lengths

The parser now recognizes two consecutive Station labels without a dash, and numeric Chainage ranges. For example, Station 1+215 Station 2+065 gives 850 meters; Chainage 1287 - Chainage 2288 gives 1,001 meters. Rates, peer groups, high-rate rankings and the FY2026 comparison use the refreshed lengths.

Earlier-rule counts are recomputed on the same current HGAB records using the previous plus-form parser, including existing explicit-length and row-bounded FMR rules. Both runs use the same records. The current rules also withhold multiple road IDs whose station references cannot safely be combined; the withheld column makes that quality correction explicit. Every record in the download includes its earlier length status and whether this update recovered a usable length. Recognizing a range does not verify its funded physical extent; adjacent endpoints do not imply overlapping work.

Download current lengths, rates and notation-recovery evidence

Very short printed ranges · original PDF text checks

The following extracted titles agree with visible text on the cited original PDF pages. This establishes printed notation, not its intended engineering meaning or the actual funded length. In particular, sub-meter road ranges require endpoint and scope clarification before interpreting allocation rates.

Download source-check evidence and full-population counts. Tiny printed ranges may reflect source notation or source-document errors. Bridges can legitimately be shorter than roads; a short bridge range alone is not an anomaly.

Length extraction and coverage methodology

• Complete STA/Station/K/KM plus-form ranges are converted to meters. Explicit “Chainage 287 - Chainage 1287” values are meter offsets, giving a 1,000-meter range. Mixed bare/plus notation, such as Chainage 200 to Chainage 3 + 500, is incomplete: the bare endpoint may lack a kilometer component. No length or rate is calculated until its units and endpoint are verified. For repeat screening only, printed endpoints are assumed to be ordered min/max: a known maximum at another segment’s known minimum excludes positive overlap even if length remains uncertain. Overlapping numerical intervals are merged, never double-counted. Multiple named routes with unverified station references are withheld.
• A single explicitly labeled length in meters or kilometers can supply a length. Widths, isolated stations and coordinate straight-line distances do not supply a road length.
• FMR length extraction uses only the current project's row-bounded source context, including printed STA labels in the analyst's coordinate cell. Context preceding the current project title is discarded, preventing a previous project’s endpoint from becoming part of the current length. Exactly two printed STA endpoints in the remaining context can form a range. The context is retained in the downloadable data; nearby unrelated text is not searched for endpoints.
• Partial or zero-length chainage, conflicting explicit lengths and ambiguous multi-route scopes remain without a usable length. No allocation amount is used to infer missing length.
• The missing percentage is (all category occurrences minus occurrences with usable stated length) / all category occurrences. It includes both absent metadata and metadata that needs review. A separate rate count distinguishes usable lengths from permissible single-category ratios.
• Mixed road/bridge titles are separated and no single-work rate is shown. Flood, building/water PAPs and foreign-assisted schedules are outside this page's denominator. Classification uses explicit titles and PAPs; it is not a complete inventory of physical road assets.
• No rate averages are pooled across categories or PAPs. Every point identifies one allocation occurrence; rankings can change with scope, width, structures, terrain, total project components and parser corrections. High rates warrant source/quantity review, not an allegation.
• Graphs and CSV follow table filters. Grouped box plots show allocation per stated meter by official program, with quartiles and every observed value; outliers are marked, not removed. Segments under 10m and rows without usable length remain in the table/CSV but are not plotted. Source volume/page and original length evidence are included. Historical stages are not used to fill current-stage gaps.

The summary links to two separate leads: length-data coverage and allocation per stated meter. Coverage gaps are extraction observations, and ratios are scope-dependent screening proxies. Extremely short printed ranges can produce extreme rates; verify station endpoints and funded scope on original pages before interpreting them. Request bills of quantities, designs, funded segment boundaries, bridge spans and roadway widths before comparing actual costs.