DTyle.AI
Real Incidents

Real Incidents. Real Lessons. Another Layer of Safety.

Safety procedures, permits, inspections, alarms and trained people are essential. DTyle.AI adds another layer: continuous visual awareness.

Book a Demo
Existing safety system

Procedures · People · Training · Permits · Inspections · Sensors · Alarms · Emergency response

DTyle.AI

Continuous visual intelligence

Human response

Alert → Assessment → Action

DTyle.AI does not replace your safety system. It extends its eyes.

What actually happened
Conception Fire — actual incident photographActual incident photograph
Image credit: U.S. Coast Guard (public domain) — salvage responseView Image Source
Real incident

Conception Fire

2 September 2019 · Santa Cruz Island, California, USA · Passenger vessel

01 — What happened?

A fire broke out at night aboard the dive boat Conception while anchored. 34 people died.

02 — Public findings
  • NTSB found the vessel had no roving night watch as required.
  • NTSB cited the lack of smoke detection in the salon area among the contributing factors.
03 — Safeguards that existed

Crew · Smoke detectors in bunk room · Regulatory night-watch requirement

Read the Original Report
NTSB investigation DCA19MM047
Where AI could potentially add another layer
Illustrative DTyle.AI camera scenario — not historical footageIllustrative DTyle.AI scenario
DTYLE.AI SAFETY ALERT
SMOKE DETECTED
LOCATION: SALON / GALLEY
CAMERA: CAM-03 · 03:14:08
STATUS: ACTION REQUIRED

Illustrative DTyle.AI analysis — not historical footage.

04 — Where was the visual opportunity?
  • Visible smoke or flame in common areas at night
  • Absence of crew movement during night hours
05 — Potential DTyle.AI intervention

Where cameras covered the salon and galley, a camera-based AI system could potentially have detected visible smoke and raised an alert to the wheelhouse and crew.

06 — Human response

Alert → operator verifies live view → crew / responders act under existing procedures.

07 — What DTyle.AI cannot do
  • Cannot replace certified fire detection
  • Requires camera coverage and adequate low-light performance
What actually happened
Deepwater Horizon / Macondo — actual incident photographActual incident photograph
Image credit: U.S. Coast Guard (public domain)View Image Source
Real incident

Deepwater Horizon / Macondo

20 April 2010 · Gulf of Mexico, USA · Offshore drilling

01 — What happened?

A blowout of the Macondo well led to explosions and fire on the Deepwater Horizon drilling unit. 11 workers died and the unit sank.

02 — Public findings
  • The USCG and joint investigations examined well control, gas detection, alarm and emergency systems.
  • Hydrocarbons reached the rig floor and ignited.
03 — Safeguards that existed

Blowout preventer · Gas detection · Alarms · Emergency shutdown · Trained crew

Read the Original Report
U.S. Coast Guard investigation hub
Where AI could potentially add another layer
Illustrative DTyle.AI camera scenario — not historical footageIllustrative DTyle.AI scenario
DTYLE.AI SAFETY ALERT
ABNORMAL DISCHARGE
LOCATION: DRILL FLOOR
CAMERA: CAM-12 · 21:41:36
STATUS: ACTION REQUIRED

Illustrative DTyle.AI analysis — not historical footage.

04 — Where was the visual opportunity?
  • Visible fluid or mud discharge on the drill floor
  • Visible gas plume / haze
  • Personnel location during emergency
05 — Potential DTyle.AI intervention

Where visible and within camera coverage, an additional visual intelligence layer could potentially have flagged abnormal discharge on the drill floor and supported muster and headcount.

06 — Human response

Alert → operator verifies live view → crew / responders act under existing procedures.

07 — What DTyle.AI cannot do
  • Cannot sense downhole pressure or gas concentration
  • Does not replace well-control engineering
What actually happened
Fremantle Highway Fire — actual incident photographActual incident photograph
Image credit: Uberprutser / Wikimedia Commons, CC BY-SA 4.0 — vessel after the fireView Image Source
Real incident

Fremantle Highway Fire

25 July 2023 · North Sea, off Ameland, Netherlands · PCTC / Ro-Ro

01 — What happened?

Fire broke out aboard the car carrier Fremantle Highway carrying several thousand vehicles. One crew member died.

02 — Public findings
  • The Dutch Safety Board investigated the fire and the subsequent emergency assistance.
03 — Safeguards that existed

Fixed fire detection · CO₂ fire suppression · Crew fire patrols · Emergency response

Read the Original Report
Dutch Safety Board investigation
Where AI could potentially add another layer
Illustrative DTyle.AI camera scenario — not historical footageIllustrative DTyle.AI scenario
DTYLE.AI SAFETY ALERT
SMOKE DETECTED
LOCATION: VEHICLE DECK 7
CAMERA: CAM-07 · 23:37:21
STATUS: ACTION REQUIRED

Illustrative DTyle.AI analysis — not historical footage.

04 — Where was the visual opportunity?
  • Visible smoke on vehicle decks
  • Early visible flame between vehicles
05 — Potential DTyle.AI intervention

A camera-based AI system could potentially have detected visible smoke on a monitored vehicle deck and provided earlier visual awareness to the bridge.

06 — Human response

Alert → operator verifies live view → crew / responders act under existing procedures.

07 — What DTyle.AI cannot do
  • Cannot detect internal battery thermal runaway before visible signs
  • Deck coverage between vehicles is limited
What actually happened

Official incident photographs are published by the investigating authority and are not reproduced here without permission.

View photographs at source
Real incident

Grande Costa D'Avorio Fire

5 July 2023 · Port Newark, New Jersey, USA · Ro-Ro / Container vessel

01 — What happened?

A fire broke out on a vehicle deck during cargo operations in port. Two Newark firefighters died.

02 — Public findings
  • NTSB investigated the fire, which began during loading operations.
03 — Safeguards that existed

Fire detection · Fixed suppression · Stevedore procedures · Port fire services

Read the Original Report
NTSB investigation DCA23FM039
Where AI could potentially add another layer
Illustrative DTyle.AI camera scenario — not historical footageIllustrative DTyle.AI scenario
DTYLE.AI SAFETY ALERT
FLAME DETECTED
LOCATION: DECK 10 — LOADING
CAMERA: CAM-21 · 21:28:44
STATUS: ACTION REQUIRED

Illustrative DTyle.AI analysis — not historical footage.

04 — Where was the visual opportunity?
  • Visible smoke or flame on the vehicle deck during loading
  • Personnel and vehicle movement in the affected area
05 — Potential DTyle.AI intervention

Where cameras covered loading decks, AI could potentially have supported earlier visual awareness of smoke and personnel locations for responders.

06 — Human response

Alert → operator verifies live view → crew / responders act under existing procedures.

07 — What DTyle.AI cannot do
  • Does not replace certified fire detection or firefighting systems
What actually happened

Official incident photographs are published by the investigating authority and are not reproduced here without permission.

View photographs at source
Real incident

X-Press Pearl Fire

May 2021 · Off Colombo, Sri Lanka · Container vessel

01 — What happened?

A fire aboard the container ship X-Press Pearl led to the loss of the vessel and significant marine pollution.

02 — Public findings
  • The IMO described it as a serious incident with major environmental consequences.
03 — Safeguards that existed

Dangerous goods declarations · Fire detection · Crew response

Read the Original Report
International Maritime Organization
Where AI could potentially add another layer
Illustrative DTyle.AI camera scenario — not historical footageIllustrative DTyle.AI scenario
DTYLE.AI SAFETY ALERT
SMOKE DETECTED
LOCATION: BAY 02 — DECK STACK
CAMERA: CAM-04 · 14:05:10
STATUS: ACTION REQUIRED

Illustrative DTyle.AI analysis — not historical footage.

04 — Where was the visual opportunity?
  • Visible smoke from a container stack
  • Visible leak or residue on deck
05 — Potential DTyle.AI intervention

Where visible, a camera-based AI system could potentially have flagged smoke from the container stack to support earlier assessment.

06 — Human response

Alert → operator verifies live view → crew / responders act under existing procedures.

07 — What DTyle.AI cannot do
  • Cannot analyse chemical composition of cargo
  • Below-deck holds may lack camera coverage
What actually happened

Official incident photographs are published by the investigating authority and are not reproduced here without permission.

View photographs at source
Real incident

MSC Flaminia Fire & Explosion

14 July 2012 · North Atlantic · Container vessel

01 — What happened?

An explosion and fire in a cargo hold of the container ship MSC Flaminia. Three crew members died.

02 — Public findings
  • The German Federal Bureau of Maritime Casualty Investigation (BSU) published an investigation report.
03 — Safeguards that existed

Smoke detection · CO₂ system · Crew response

Read the Original Report
BSU investigation report 255/12 (PDF)
Where AI could potentially add another layer
Illustrative DTyle.AI camera scenario — not historical footageIllustrative DTyle.AI scenario
DTYLE.AI SAFETY ALERT
SMOKE DETECTED
LOCATION: HOLD 4 — HATCH
CAMERA: CAM-09 · 08:02:57
STATUS: ACTION REQUIRED

Illustrative DTyle.AI analysis — not historical footage.

04 — Where was the visual opportunity?
  • Visible smoke from hatch covers or ventilation
  • Crew positions during emergency response
05 — Potential DTyle.AI intervention

An additional visual intelligence layer could potentially support response by showing visible smoke on deck and the location of crew.

06 — Human response

Alert → operator verifies live view → crew / responders act under existing procedures.

07 — What DTyle.AI cannot do
  • Cannot see inside closed holds or detect chemical decomposition
What actually happened

Official incident photographs are published by the investigating authority and are not reproduced here without permission.

View photographs at source
Real incident

Piper Alpha

6 July 1988 · North Sea, UK · Offshore oil & gas

01 — What happened?

Explosions and fire destroyed the Piper Alpha platform. 167 people died.

02 — Public findings
  • The public inquiry examined permit-to-work, communication and emergency response failings.
03 — Safeguards that existed

Permit-to-work system · Gas detection · Fire protection · Emergency procedures

Read the Original Report
UK HSE — Public Inquiry
Where AI could potentially add another layer
Illustrative DTyle.AI camera scenario — not historical footageIllustrative DTyle.AI scenario
DTYLE.AI SAFETY ALERT
FIRE / SMOKE
LOCATION: PRODUCTION MODULE
CAMERA: CAM-31 · 22:00:14
STATUS: ACTION REQUIRED

Illustrative DTyle.AI analysis — not historical footage.

04 — Where was the visual opportunity?
  • Visible work activity on equipment under maintenance
  • Visible smoke or flame
  • Personnel muster
05 — Potential DTyle.AI intervention

DTyle.AI analysis identifies a potential visual opportunity: cameras could potentially support awareness of work status and personnel locations during escalation.

06 — Human response

Alert → operator verifies live view → crew / responders act under existing procedures.

07 — What DTyle.AI cannot do
  • Cannot verify isolation status or replace a permit-to-work system
What actually happened

Official incident photographs are published by the investigating authority and are not reproduced here without permission.

View photographs at source
Real incident

SEACOR POWER

13 April 2021 · Off Port Fourchon, Louisiana, USA · Offshore / Liftboat

01 — What happened?

The liftboat SEACOR POWER capsized in severe weather. 13 people died.

02 — Public findings
  • A USCG Marine Board of Investigation examined the capsizing and search-and-rescue response.
03 — Safeguards that existed

Weather forecasting · Stability procedures · Lifesaving equipment · SAR

Read the Original Report
U.S. Coast Guard investigation
Where AI could potentially add another layer
Illustrative DTyle.AI camera scenario — not historical footageIllustrative DTyle.AI scenario
DTYLE.AI SAFETY ALERT
PERSON IN WATER
LOCATION: PORT SIDE
CAMERA: CAM-02 · 15:39:02
STATUS: ACTION REQUIRED

Illustrative DTyle.AI analysis — not historical footage.

04 — Where was the visual opportunity?
  • Visible persons in water
  • Visible deck conditions and crew locations
05 — Potential DTyle.AI intervention

Where camera coverage exists, AI could potentially support the response by detecting visible persons in the water.

06 — Human response

Alert → operator verifies live view → crew / responders act under existing procedures.

07 — What DTyle.AI cannot do
  • Cannot predict capsizing or replace stability calculations
  • Heavy weather limits visibility
What actually happened

Official incident photographs are published by the investigating authority and are not reproduced here without permission.

View photographs at source
Real incident

Al Hamriyah Port Fire

2025 · Sharjah, UAE · Petrochemical / Port

01 — What happened?

A fire at Al Hamriyah Port was contained by emergency services; injuries were reported by official UAE sources.

02 — Public findings
  • Public investigation findings currently available do not establish a cause. No cause is stated here.
03 — Safeguards that existed

Civil defence response · Port emergency procedures

Read the Original Report
Emirates News Agency (WAM)
Where AI could potentially add another layer
Illustrative DTyle.AI camera scenario — not historical footageIllustrative DTyle.AI scenario
DTYLE.AI SAFETY ALERT
SMOKE DETECTED
LOCATION: YARD B
CAMERA: CAM-15 · 19:12:40
STATUS: ACTION REQUIRED

Illustrative DTyle.AI analysis — not historical footage.

04 — Where was the visual opportunity?
  • Visible smoke or flame in the yard
  • Personnel near the affected area
05 — Potential DTyle.AI intervention

A camera-based AI system could potentially provide earlier visual awareness of smoke in a monitored yard.

06 — Human response

Alert → operator verifies live view → crew / responders act under existing procedures.

07 — What DTyle.AI cannot do
  • Cannot establish cause
  • Does not replace gas or heat detection

What AI vision can see — and what it can't

DTyle.AI can sit within a multi-sensor safety architecture. Computer vision does not solve every hazard.

Visual AI can potentially see
PersonSmokeFlamePPEUnsafe activityRestricted-zone entryVehicle movementFallsVisible leaksVisible obstructionsVisible hot workPerson in waterCrowding
Other sensors may be required
GasVOCPressureInternal temperatureChemical compositionBattery internal temperatureTank atmosphereDownhole conditionsMachine vibrationElectrical faults

Every incident leaves a lesson.

The objective isn't to replace the safety systems that already exist. It's to add another set of eyes — continuously, consistently and across the areas your cameras cover.

See Risk. Detect Events. Create Awareness. Support Action.

Request a Demo Discuss a PilotExplore Use Cases