LIVEdataset aec-bench@releasetasks 552models 18last submission … · built …
civilwith-tool

Ssd On Grade

Computes stopping sight distance (SSD) on graded road segments as the sum of reaction distance d_r = V*t_r/3.6 and braking distance d_b = V^2/(254*(f+g)), using speed-dependent longitudinal friction coefficients from AGRD Part 3 Table 5.5. Accounts for uphill and downhill grades to verify geometric design adequacy for safe vehicle stopping.

with-tool: The model is given an executable Python calculator script.

Standards

AGRD Part 3 §5

Tags

civildeterministicgraderoadssight-distancestopping

How this task is generated

One template produces many comparable benchmark tasks while keeping the scoring contract fixed.

  1. 01

    Template

    The reusable contract shown on this page.

  2. 02

    Scenario

    An archetype and site context are sampled.

  3. 03

    Difficulty tier

    Inputs may be hidden at harder tiers.

  4. 04

    Task prompt

    The model responds with the declared outputs.

Parameters

Inputs the model receives, and the outputs it is scored on.

Inputs

3 inputs

Always given

Included directly in every task prompt.

2

  • Design speed

    design_speed_km_h

    Design speed V

    40 – 130 km/h

  • Grade pct

    grade_pct

    Longitudinal grade (positive = uphill, negative = downhill)

    -10 – 10 %

Hidden at higher difficulty

Visible in easier tasks and withheld in one or more harder tiers.

1

  • Reaction time

    reaction_time_s

    Driver reaction time t_r

    Hidden at hard difficulty.

    1.5 – 2.5 s

Scored outputs

3 outputs

Reaction distance

reaction_distance_m

Reaction distance component d_r (m)

Scores if within ±3% of the reference value.

Braking distance

braking_distance_m

Braking distance component d_b (m)

Scores if within ±3% of the reference value.

Stopping sight distance

stopping_sight_distance_m

Total stopping sight distance SSD (m)

Scores if within ±3% of the reference value.

Difficulty

Each template is sampled at three tiers. Harder tiers may hide inputs, forcing the model to infer them from the scenario description.

easy

All inputs given

All parameters given, flat to gentle grade on a suburban arterial

medium

All inputs given

All parameters given, any road type including steep grades

hard

Some inputs hidden

Reaction time hidden — agent must infer t_r from road type and driver alertness context

Hidden inputs

  • Reaction time sreaction_time_s

Prompt replacement text

The site is a {{ archetype.description }} ({{ archetype.site_context }}). Assume an appropriate driver reaction time for this road environment.

Task bundle

The exact instruction and parameter contract used to generate this task, pinned to the published library source.

/workspace

  • instruction.md
  • ssd-on-grade_calc.py

Teal lines show Jinja input conditions, not task visibility policy. A line renders only when that input or tool is visible.

1You are a senior civil engineer specializing in road geometry and stopping sight distance analysis.2 3## Problem4 5Calculate the required stopping sight distance (SSD) for a road segment on a longitudinal grade.6 7## Given8 9| Parameter | Value | Unit |10|-----------|-------|------|11| Design speed (V) | {{ design_speed_km_h }} | km/h |12| Longitudinal grade | {{ grade_pct }} | % |13{% if reaction_time_s is defined %}14| Reaction time (t_r) | {{ reaction_time_s }} | s |15{% endif %}16{% if archetype_description is defined %}17 18### Site Conditions19 20{{ archetype_description }}21{% endif %}22 23{% if tool_available %}24## Available Tool25 26A stopping sight distance calculation tool is available at `/workspace/{{ meta.name }}_calc.py`. Run it with:27 28```bash29python3 /workspace/{{ meta.name }}_calc.py --help30```31 32You may use this tool to verify your calculations or compute values directly.33{% endif %}34 35## Required36 37Calculate the following:38 391. Reaction distance component d_r (m)402. Braking distance component d_b (m)413. Total stopping sight distance SSD (m)42 43## Applicable Standards44 45- Austroads Guide to Road Design Part 3 (AGRD Part 3 §5)46 47## Constraints48 49- No internet access is available. Work from engineering knowledge and the provided tool.50- The friction coefficient f depends on design speed per AGRD Table 5.5. Use the following lookup:51 - 40 km/h → f = 0.52, 50 → 0.48, 60 → 0.45, 70 → 0.43, 80 → 0.40, 90 → 0.38, 100 → 0.36, 110 → 0.34, 120 → 0.32, 130 → 0.2952 - Interpolate linearly for intermediate speeds53- Reaction distance: d_r = V × t_r / 3.6 (where V in km/h, t_r in seconds)54- Braking distance: d_b = V² / (254 × (f + g)) (where g = grade / 100, positive = uphill)55 - Sign convention: positive grade means uphill travel (grade assists braking); negative grade means downhill travel (grade hinders braking)56- Total SSD = d_r + d_b57{% if reaction_time_s is not defined %}58- Reaction time is not provided; select an appropriate value based on the road environment and expected driver alertness. AGRD recommends 1.5 s (alert) to 2.5 s (relaxed).59{% endif %}60 61## Output Format62 63Show your step-by-step working in Markdown. At the end of your solution, include a JSON block with your final answers in exactly this format:64 65```json66{67 "reaction_distance_m": <numeric_value>,68 "braking_distance_m": <numeric_value>,69 "stopping_sight_distance_m": <numeric_value>70}71```72 73Write your complete solution to `/workspace/output.md`.74

Scenario archetypes

Each generated task is drawn from one of these realistic scenario bands.

Site contexts ground each scenario in a real locale the model can use to infer hidden values.

Urban local street

urban_local_street

Low-speed urban local street with pedestrian activity and on-street parking

sydney-residential-streetmelbourne-local-roadbrisbane-suburban-street
Parameter ranges
design_speed_km_h
40 – 60
reaction_time_s
2 – 2.5
grade_pct
-5 – 5

Suburban arterial

suburban_arterial

Suburban arterial road with moderate speeds and intersection access

sydney-pennant-hills-roadmelbourne-springvale-roadbrisbane-waterworks-road
Parameter ranges
design_speed_km_h
60 – 80
reaction_time_s
1.5 – 2.5
grade_pct
-6 – 6

Rural highway

rural_highway

Rural two-lane highway through open terrain with higher operating speeds

bruce-highway-qldprinces-highway-vicpacific-highway-nsw
Parameter ranges
design_speed_km_h
80 – 110
reaction_time_s
1.5 – 2.5
grade_pct
-8 – 8

Mountain road

mountain_road

Winding mountain road with steep grades and limited sight lines

great-ocean-road-vicbells-line-of-road-nswgillies-highway-qld
Parameter ranges
design_speed_km_h
60 – 100
reaction_time_s
2 – 2.5
grade_pct
-10 – 10

Example task

sydney-pennant-hills-road-suburban-arterial-preview — hard difficulty, some inputs hidden.

Suburban arterial road with moderate speeds and intersection access. sydney-pennant-hills-road. Required outputs: reaction_distance_m, braking_distance_m, stopping_sight_distance_m

The model sees

Scenario context and visible inputs.

design_speed_km_h
60 to 80 km/h
grade_pct
-6 to 6 %

Executable tool: ssd-on-grade_calc.py

The model must infer

Inputs withheld at this difficulty.

  • Reaction time s

    reaction_time_s

Stand-in text in the prompt

The site is a {{ archetype.description }} ({{ archetype.site_context }}). Assume an appropriate driver reaction time for this road environment.

The model must produce

The scored JSON answer schema.

{
  "reaction_distance_m": <number>,
  "braking_distance_m": <number>,
  "stopping_sight_distance_m": <number>
}
  • reaction_distance_m · scored within ±3%
  • braking_distance_m · scored within ±3%
  • stopping_sight_distance_m · scored within ±3%