Practice with national-level exam (FACT, FACT Plus, NET, CUET, etc.) mocks, learn from structured notes, and get your doubts solved in one place.
Scene-preservation in moving-traffic environments, skid mark physics with the S = √(30·d·μ·n) formula, collision typology, vehicle examination protocols, and the Indian hit-and-run workflow under BNS 106 and BNSS.
Accident investigation is the working name for the forensic-reconstruction discipline that turns a chaotic crash scene into a defensible answer about who did what, how fast, and in what sequence. It sits in Module 4 because the evidence is pattern evidence at the scale of an entire roadway: skid marks, yaw arcs, gouge tracks, debris fans, paint smears on victim clothing, glass embedded in a tyre tread. The investigator has minutes to document a dynamic environment before traffic flow destroys the geometry, and then weeks to defend the conclusions in a magistrate's court under BNS 106 (causing death by negligence) and the time-bound investigation regime of BNSS 2023.
Hit-and-run is the harder subset. The vehicle is gone, the driver is gone, and the only material left is what stayed on the road and what stuck to the victim's clothes. Paint chips, glass fragments, tyre track impressions, headlight filament evidence, and the EDR module that sits behind the dashboard of every car sold in India after 2024. The reconstruction question splits in two: where was the vehicle, and how fast was it moving. The first is geometry. The second is the drag-factor formula, applied to skid marks that the SOCO photographed and measured before the road reopened.
The road is the evidence, and the road wants to reopen.
A crash scene differs from every other scene in this syllabus on one axis: time pressure to clear it. A homicide in a bedroom can be cordoned for 24 hours. A fatal collision on NH-48 outside Mumbai loses control of the lane in roughly 45 minutes. Skid marks degrade with the next monsoon shower, gouge marks fill with rubber dust by the next truck, and debris drifts every time a vehicle passes. The whole reconstruction depends on documentation made within the first hour.
The first-officer protocol is therefore aggressive about geometry capture before evidence collection. Photographs and a measured sketch come first; physical recovery comes second. Tyre-mark lengths are paint-marked in chalk or spray before the road is washed down. The point of rest of every vehicle and every body is fixed by triangulating from two permanent reference points (a kilometre post, a culvert edge, a survey marker on the carriageway).
A working checklist for the first 60 minutes:
Six patterns, six different physics arguments.
The collision type sets the entire reconstruction approach. The damage signature on each vehicle, the resting positions, and the casualty pattern all flow from the geometry of impact. Indian FACT examiners ask candidates to classify a scene from a photograph in under thirty seconds, so the typology has to be reflex.
| Collision type | Geometry | Signature damage | Reconstruction priority |
|---|---|---|---|
| Head-on | Two vehicles approach on a near-180° line | Front-end crush on both, hood deformation, steering column displacement | Combined closing speed, lane-position at impact |
| Rear-end | Following vehicle strikes leading vehicle from behind | Front of follower, rear of leader; whiplash in lead occupants | Following distance, brake-reaction time, lead vehicle's deceleration profile |
| T-bone (right-angle) | Side of one vehicle struck by front of another at ~90° | Intrusion damage on the struck vehicle's flank, B-pillar deformation | Signal-phase data, right-of-way determination, speed of striking vehicle |
| Glancing/sideswipe |
Forty feet of black rubber tells you the minimum speed at the start of the skid. Not the impact speed.
The classic skid-mark speed formula in Indian and US accident-reconstruction practice is:
S = √(30 · d · μ · n)
Where S is speed in miles per hour, d is the skid distance in feet, μ is the drag factor (effective coefficient of friction) and n is the braking efficiency, a fraction between 0 and 1 that accounts for how many of the four wheels were locking. The metric equivalent, with S in km/h and d in metres, is S = √(254 · d · μ · n), which most Indian SFSL reports now use as the primary form.
A worked feel for the number: a 30-metre skid on dry asphalt (μ = 0.75) with all four wheels locked (n = 1) gives S = √(254 · 30 · 0.75 · 1) = √5715 ≈ 76 km/h. That is the minimum speed at the start of the skid. The impact speed will be lower, sometimes much lower, because the vehicle continued to decelerate over the residual distance to impact.
The drag-factor table every candidate should memorise:
Braking efficiency n is the variable candidates skip and lose marks on. If only the front wheels lock, n drops to roughly 0.65 because the rear wheels still contribute residual rolling-tyre friction that the locked-wheel formula does not capture. If one wheel locks and three roll, n falls further. A modern ABS-equipped vehicle, by design, never produces a fully locked-wheel skid; it produces a series of intermittent scuff marks, and the formula above does not apply at all. ABS cases need an EDR pull instead.
Five different stories, five different physics arguments.
A reconstruction stands or falls on whether the investigator correctly identifies what each mark on the road is. Conflating a yaw mark with a braking mark is the most common error in Indian SFSL hit-and-run reports, and one that defence counsel use to break the prosecution's speed estimate.
Paint, glass, filaments, tyres, undercarriage, EDR.
Once a candidate vehicle is recovered, whether at the scene or weeks later through ANPR work as set out in Speed Detection Devices, the examination protocol has six anchor checks. Each one is documented separately in the SFSL report.
From the scene to the SFSL bench to the magistrate, on a 90-day clock.
A hit-and-run case in India runs through a defined sequence: scene response, victim recovery, evidence collection, vehicle search via ANPR plus FASTag data, vehicle examination at SFSL, paint and glass comparison, and the file submission within the BNSS time-bound investigation window. BNS 106 prescribes up to five years for causing death by negligence in a hit-and-run where the driver does not report; BNSS 2023 requires the investigation to be completed within 90 days for offences punishable up to 10 years.
The high-yield evidence carriers in a hit-and-run, in priority order:
The common errors that recur in SFSL hit-and-run files reviewed by NFSU faculty: treating skid length as full stopping distance (under-estimates speed); failing to record road gradient (a 6 % downhill grade changes effective μ); conflating yaw with skid in single-vehicle loss-of-control cases (wrong formula, wrong answer); and reporting the EDR speed without authenticating the 65B/63 certificate at the time of seizure.
A 25-metre continuous skid mark on dry asphalt with all four wheels locked. Drag factor μ = 0.75, braking efficiency n = 1. Using S = √(254·d·μ·n) in metric units, the speed at the start of the skid is closest to:
| Parallel paths, lateral contact |
| Long parallel scrapes on side panels, mirror sheared off, paint transfer |
| Lane discipline failure, paint-layer matching for hit-and-run |
| Rollover | Vehicle pitches or rolls onto its side or roof | Roof crush, glass collapse, side and roof scrape marks on pavement | Trip mechanism (curb, soft shoulder, yaw exceeding limit) |
| Pedestrian-vs-vehicle | Vehicle strikes a person on or off the carriageway | Bumper-height bruise on victim, windscreen damage, throw distance | Vehicle speed from throw equations, point of impact, evasive action |
The pedestrian collision deserves a separate paragraph because it carries the bulk of hit-and-run casework in India. The bumper-height contact bruise on the victim's leg gives an approximate height of the striking vehicle's bumper at the moment of impact, which narrows the candidate-vehicle list. The throw distance from impact to point of rest, combined with the victim's mass and stance, allows a rough speed estimate using one of the Searle or Wood equations. Glass embedment in the victim's clothing, paint smear at the contact point, and tyre tread on the legs or torso of a run-over victim are the high-yield items.
Two practical examples worth carrying into the viva: a head-on on a divided highway is almost always a wrong-way driver or a lane-departure event, and the reconstruction will hinge on which vehicle crossed the centre line first. A T-bone at an urban intersection is almost always a signal-violation case, and the reconstruction will hinge on the signal-phase log if the junction is CCTV-monitored.
The yaw mark is the high-value target in a single-vehicle loss-of-control case, and one of the few situations where speed can be calculated without any skid. The vehicle is mid-corner, the driver's input has exceeded available grip, and the car slides outward in an arc. Measure the chord of the arc, measure the middle ordinate, compute the radius, plug into S = √(127·R·μ), and you have an estimate that is independent of braking distance.
In a 2025 South Mumbai hit-and-run handled by Maharashtra SFSL, the driver claimed he was travelling at 40 km/h when a pedestrian stepped onto the carriageway. The EDR recovered from the BMW showed 96 km/h three seconds before impact, with no brake application until 0.4 s before contact. The discrepancy between the driver's statement and the recorded data was the centre of the prosecution case under BNS 106. EDR evidence has become the most contested category in Indian accident-reconstruction trials.