System role, port scope, and accountability boundary
A port operations fatigue risk system is a management mechanism through which a port authority, terminal operator, and relevant employers connect rosters, staffing, operating windows, task hazards, fatigue reports, appropriate monitoring, field controls, and continual improvement to reduce fatigue-related error, injury, and post-shift travel risk. It belongs to WSIC 4.2.4. It may cover container, dry bulk, liquid bulk, roll-on/roll-off, passenger, and multipurpose terminals, but it must state the facilities, cargoes, roles, contracting arrangements, and jurisdictions in scope.
Shore personnel, longshore workers, terminal vehicle and equipment operators, line handlers, tally staff, and maintainers are governed differently from ship crews, pilots, road carriers, railway workers, and temporary visitors. The system should define the ship-shore boundary, handoff points, and control held by each party. Crew rest rules cannot simply become shore-worker roster rules, and a third party's entry into the port does not remove shared risk duties. The system is not merely an hours calculator, camera, wearable, or medical fitness tool and cannot automatically clear or discipline a person or change law, collective agreements, and safe procedures.
Operation chain, task hazards, and complete exposure
The risk map should cover berthing, gangway and access setup, line handling, quay and ship cranes, work in holds, lashing, yard cranes, straddle carriers, reach stackers, forklifts, terminal tractors, yard and gate work, reefer connection, dangerous goods, bunkering, maintenance, cleaning, and emergencies. Work at height, suspended loads, crush zones, ship-shore gaps, vehicle blind spots, oxygen deficiency or toxic atmosphere in holds, noise, heat or cold, wind, rain, and poor night lighting magnify fatigue consequences. One common score cannot decide continuation across all tasks.
The timeline should begin with call-in and commuting and include terminal security, pre-shift meetings, changing, equipment collection and inspection, handover, vessel delay, actual production, equipment and weather stops, cargo changes, overtime, close-out, exit, and a safe journey home. Planned and actual shifts, standby, recall, secondary work, training, breaks, and interruptions require separate records, with time zone, daylight-saving status, vessel-schedule revision, source, and missing-data reason. Waiting is not necessarily recovery, and a short stoppage does not automatically create a sleep opportunity. Off-terminal queues, transfers across facilities, and post-shift driving remain part of the fatigue-risk chain.
Rosters, staffing, and ship-shore coordination
Roster design should combine arrival variability, tide and pilot windows, cargo volume, equipment availability, skill mix, consecutive nights, rotation direction, continuity of rest, and commuting instead of optimizing only predicted throughput. Every role needs qualifications, equipment authorization, work area, supervision, and relief relationships. Casual, agency, and contractor workers need equivalent training, reporting, and protection. Relief must be qualified, adequately recovered, and able to reach the post, and moving that person must not deprive another work face of necessary lookout, control, or emergency capability.
If a vessel arrives early or late, shifts berth, meets severe weather, suffers an equipment failure, changes cargo, or lacks staff, later work and rosters should be reassessed with actual information rather than inheriting the original green status. The vessel, terminal, carrier, railway, pilot service, and control room need sufficient exchange of operating stage, delay, dangerous cargo, people, and equipment state, while each retains its legal responsibilities. Any exception that extends a shift or exceeds an internal boundary needs explicit authorization, task-level assessment, added controls, and later recovery. Repeated exceptions are not a normal business model.
Layered controls, alerts, and on-shift response
Controls should first reduce exposure by providing adequate staffing and reserve skills, improving vessel and labor forecasts, limiting consecutive nights and short-notice calls, protecting rest and handover, deferring noncritical maintenance or training, and preventing piece rates, demurrage, or truck-turn pressure from driving unsafe speed. Engineering controls include separation of people and vehicles, lighting, clear traffic routes, cab ergonomics, anticollision and limiting devices, reliable communication, weather protection, and suitable recovery areas. Fatigue software cannot replace any of these controls.
An alert needs a defined recipient, acknowledgement, verification, stop, relief, and closure path. Mobile-equipment operators should stop in a suitable safe place. A quay crane or lifting operation requires stabilization of the load and transfer of control before relief, while line handling and hold work require role-specific withdrawal and attendance. A fatigued worker should not complete a long questionnaire or repeatedly acknowledge prompts while operating. Responses may include qualified relief, task rotation, greater supervision, deferral of high-consequence work, recovery, transport adjustment, or a safe trip home. Caffeine, cold air, music, and brief activity do not prove that extended high-risk work can continue.
Monitoring technology, worker reports, and health boundary
Biomathematical models, subjective scales, reaction time, vehicle or crane controls, operator-facing cameras, and wearables can supply supporting evidence if inputs, operating domain, thresholds, false alerts and misses, unavailable states, and version are declared. Lighting, glasses, personal protective equipment, posture, vibration, roads, loads, and automation can alter measurements. No data, an obscured camera, transmission failure, or no report cannot be treated as low risk. A score must not override field observation, actual hours, or self-report and cannot independently determine discipline or accident responsibility.
Workers who believe they cannot operate safely need a reporting and protection path that a “normal” algorithm cannot reject, and peers and supervisors should be able to raise an immediate concern. The operator should distinguish good-faith reports, ordinary error, recklessness, and deliberate violation and explain confidentiality, feedback, and discipline. A decline in reporting is not evidence of less risk. Persistent uncontrollable sleepiness, severe insomnia, possible sleep-disordered breathing, pain, or concern about sedating medicine belongs in a confidential occupational-health pathway. The system cannot diagnose from body size, cameras, or questionnaires, advise medication cessation, or disclose illness broadly to production staff.
Validation, investigation, and change control
Before launch, historical replay and shadow operation should cover nights, consecutive vessel calls, schedule delays, tidal windows, severe weather, equipment failure, cargo peaks, staffing gaps, contractor crews, and communications loss. Validation should test not only rules and algorithms but whether qualified relief, parking, load stabilization, handover, and safe transport are genuinely available. Measures can normalize fatigue concerns, realized rest, temporary relief, alert response, false alerts and misses, unavailable time, near misses, and injuries by work hours, equipment hours, container moves, trips, or task exposure. Throughput, logins, or fewer reports alone cannot establish safety improvement.
Accident and near-miss investigation should reconstruct actual rosters, time awake and sleep opportunity, commuting, vessel changes, task environment, equipment and load state, ship-shore communication, reports, alerts, human decisions, and model version. It should examine interacting organizational barriers rather than blame the last operator. A new terminal, cargo, vessel type, crane or automation system, yard layout, traffic route, shift pattern, contractor, collective agreement, interface, or algorithm requires impact assessment and renewed validation when needed. Recurrent overtime, overrides, or inability to stop safely should feed back to vessel planning, staffing, and site design.
Data governance, cybersecurity, and continuity
Combined roster, location, video, biometric, health, and performance data may affect employment and labor relations. Collection should be minimized to safety purposes, with clear roles for employers, terminals, contractors, vessels, occupational health, and regulators. Personal data should not default to productivity ranking, unrelated discipline, insurance, or commercial profiling. Medical information and operational decisions need role separation. Workers should have access, correction, and appeal rights; cross-company sharing should be limited to immediate safety need, and reports from small teams require protection against re-identification by role or event.
Procurement should test low bandwidth, wireless dead zones, offline equipment, shared terminals, shift handover, clock drift, and cyberattack, and define system isolation, access revocation, vulnerability and supply-chain management, rule and model updates, audit, export, rollback, and deletion at termination. A monitoring platform must not directly control lifting, vehicle braking, or another safety-critical function without an independent safety case. During power, network, or supplier outage, stale data need explicit warning and approved manual rosters, communications, risk checks, and reporting. A terminal cannot outsource safety and compliance accountability to software.