Safety, risk and regulatory compliance in transport and logistics
This domain tests structured risk reasoning and your command of the Hong Kong statutory framework covering roads, buses, ferries, ports, airports and cargo handling.
Risk framework
Risk is a function of likelihood and severity; ALARP requires that risk be reduced until further reduction is grossly disproportionate in cost. Use the hierarchy of controls (eliminate, substitute, engineer, administrate, PPE) and a bow-tie to separate preventive barriers from mitigating recovery measures. The Haddon matrix sorts countermeasures by pre-event, event and post-event, across human, vehicle and environment.
Road safety engineering
Road Safety Audit covers five stages from feasibility to post-opening, conducted by an independent team. Network screening ranks sites by accident rate per 100 million vehicle-kilometres, not raw counts, and empirical Bayes estimation shrinks random fluctuation (regression to the mean) before a site is treated.
Worked example
A 3.2 km rural road with AADT 60,000 recorded 45 injury accidents in a year.
- Vehicle-km = 60,000 × 365 × 3.2 = 70.08 million.
- Rate = 45/70.08 = 0.642 per million veh-km = 64.2 per 100 million veh-km.
If the comparator network average is 45 per 100 million veh-km, the site is about 43% above average and a candidate for investigation - but confirm with several years of data and check exposure changes (a new development may have raised AADT) before concluding.
Regulatory map
- Road Traffic Ordinance (Cap 374) and its regulations: vehicle standards, traffic control, offences.
- Public Bus Services Ordinance (Cap 230); Ferry Services Ordinance (Cap 104).
- Airport Authority Ordinance (Cap 483) and airside safety rules; ISAGO for ground handlers.
- Shipping and Port Control Ordinance (Cap 313); Dangerous Goods Ordinance (Cap 295) and IMDG Code for sea freight.
- Factories and Industrial Undertakings Ordinance (Cap 59) for cargo-handling safety; EIA Ordinance (Cap 499) for major projects.
Traps that cost marks
- Comparing accident counts without exposure (traffic volume, length, years).
- Ignoring regression to the mean when selecting blackspots.
- Citing ALARP without any cost–benefit reasoning.
- Confusing airside, landside and marine regulatory regimes in one answer.