long op• Data kinds: scalar × scalar → table
• Call: import fullseye as fs; fs.ledger.stop_line_plan(v: 'float', dist_to_line: 'float', theta: 'float' = 0.0, params: 'Optional[dict]' = None, t0: 'float' = 0.0, margin: 'float' = 0.5, A1: 'float' = 1.5, A2: 'float' = 2.5, split: 'float' = 0.5) -> 'Dict[str, object]' (to call the implementation directly, import drivelong; drivelong.stop_line_plan(v: 'float', dist_to_line: 'float', theta: 'float' = 0.0, params: 'Optional[dict]' = None, t0: 'float' = 0.0, margin: 'float' = 0.5, A1: 'float' = 1.5, A2: 'float' = 2.5, split: 'float' = 0.5) -> 'Dict[str, object]'; from the registry, opsdrive.get("stop_line_plan"))
From the moment of perception (time t0, dist_to_line from the front end to the stop line, speed v), plans a 2-stage braking that stops margin short of the stop line.
Phases: reaction (ρ, speed held) → [if needed, wait while holding speed (cruise)] → stage 1 (constant part of deceleration A1, until speed reaches split·v) → stage 2 (A2, until stopped) → hold. Braking is split into two stages to match the official notice's "not applying the brake in several steps" (improper braking operation). If it cannot stop in time, the stage-2 deceleration is re-solved in closed form (the ln formula, exact even for k > 0); if stage 1 alone covers the distance, it stops in one stage (stages = 1). If capped at the upper limit (min(a_brake_max, μ g cos θ)), saturated = True and the prediction uses the capped value. The slope is assumed to have constant θ over the interval (the stopping interval does not cross a break in the grade).
dict : phases (a list of (t_start, kind, drive, brake), kind = react / cruise / brake1 / brake2), t_breaks,
stages, predicted_gap (from the stopped front end to the stop line; negative = overshot), t_stop, saturated, theta, d_react, d_cruise, d_brake.
Raises ValueError: v ≤ 0, non-finite, split ∉ (0, 1), A1 or A2 ≤ 0.
• Sample-data catalog (download URLs / licences) — 2-D uses skimage.data (BSD/public domain) plus synthetic images; 3-D lists download URLs for real data sources (Stanford, PDS, …).
• Operator provenance and references — the sources of the research/methods this op family came from.
• The canonical algorithm (author, year) and its uses are named in the family usage guide above.
• poc_driving_longitudinal — py -3.11 examples/poc_driving_longitudinal.py
• poc_driving_weather — py -3.11 examples/poc_driving_weather.py
table as input)course_layout · course_occupancy · course_contains · world_build · world_camera · world_move · lidar_scan · rss_longitudinal_same
long)long_params · road_profile · road_eval · long_simulate · long_energy_residual · stopping_distance_grade · plan_command · hill_hold_brake_min
*Provenance: drivelong.py — DRIVE operator registry. This per-op note is generated by tools/opdocs.py md (do not hand-edit).*
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.