A Goliath Crane Collision Prevention Solution That Detects Even Unregistered Obstacles in Shipyard Yards

Crane sites already have anti-collision systems. The problem is they only know pre-registered obstacles. A pipe on the rails or an aerial lift that arrived today simply doesn't exist to them. LookOut detects people, equipment, and irregular objects on the rails in 3D, with no prior data.

Oct 07, 2026

A Goliath Crane Collision Prevention Solution That Detects Even Unregistered Obstacles in Shipyard Yards

Contents

Who Is This For?A story for organizations responsible for sites where heavy loads move along fixed paths.The Problem on the Ground While a Goliath crane travels along its rails, four problems overlap.Why Were Existing Methods Insufficient?How LookOut Solves It What Changes In a Nutshell Featured Use Case Building a Yard Without Blind Spots, LookOut Is With You

Who Is This For?

A story for organizations responsible for sites where heavy loads move along fixed paths.

Heavy industry companies operating shipyards and offshore plant yards
Safety management and equipment teams that operate Goliath cranes and gantry cranes on a daily basis
Port cargo-handling companies operating container terminals
Planning and DX departments driving yard automation and smart shipbuilding
Crane manufacturers and maintenance companies, and industrial safety system integrators

They share the same conditions: the objects they handle are heavy, and a single accident can bring the entire yard's operations to a halt.

The Problem on the Ground

While a Goliath crane travels along its rails, four problems overlap.

Safety: The area beneath each crane leg and the travel rails cannot be seen from the operator's cab. The very points where contact occurs are blind spots.
Technology: The objects to be monitored are not uniform. Not only people, bicycles, forklifts, and aerial work platforms, but also wires and pipes left on the rails all count as intrusions.
Environment: The yard is outdoors, fully exposed to rain, sea winds, and seasonal temperature changes.
Decision-making: Situations that came close to contact but stopped in time go unrecorded, leaving no basis for identifying which points are dangerous.

Why Were Existing Methods Insufficient?

The root of the problem lies in what existing measures assume in order to work.

Pre-registration–based anti-collision systems make decisions based on the positions of obstacles entered in advance. Any unregistered object that enters the rails is not detected.
Worker monitoring and hand signals cannot cover the entire length of the rails at all times, and communication is delayed at night and in rain.
Camera-only monitoring relies on 2D images, making it difficult to judge the actual distance and height of objects.
Tags and wearable devices apply only to people. They cannot be used for inanimate obstructions such as wires or pipes.
The absence of records leaves no basis for setting improvement priorities.

In short, there was no solution that simultaneously met all four conditions: working without pre-registration, detecting irregular objects, operating reliably outdoors, and keeping a record.

How LookOut Solves It

LookOut reads the area around the crane in 3D using LiDAR and assesses risk through the following steps.

Detection and classification: People, vehicles, and other objects are detected and distinguished as separate 3D objects. Because LookOut's proprietary AI algorithm detects even irregular objects not included in its training list, intrusions on the rails are recognized regardless of whether they were registered in advance.
Omnidirectional monitoring: The front, rear, left, and right of the area beneath each crane leg are monitored in 360 degrees, capturing zones beyond the cab's line of sight as coordinates.
3D information capture: Distance, direction, height, speed, and movement trajectory are calculated together for each detected object. Contact risks that cannot be judged by horizontal distance alone are calculated with height included.
Tiered hazard zones: Zones divided into danger, caution, and safe are defined with precise coordinates. When work layouts or equipment positions change, only the zones are reconfigured, without moving any structures.
Assessment and integration: When an object enters a hazard zone, the system judges it as a collision risk and sends the result to the existing crane anti-collision system and brake interlock circuit. Real-time video and distance information are also provided through the cab interface.
Sensor fusion: Cameras and LiDAR are calibrated so that 3D perception results are overlaid on the video. Managers can check coordinates and the screen at a glance.
Recording and analysis: Event histories and observation data are stored automatically for later review and analysis.
De-identification: Because people are recognized through 3D coordinates rather than raw video, worker positions are detected without identifying faces.

No pre-registration required, irregular objects included, built for outdoor environments, and a full record. LookOut meets all four conditions above at once.

What Changes

When the premise of assessment changes, the nature of safety management changes.

Because detection does not depend on a registration list, equipment and materials that entered the yard for the first time that day are also detected.
The space beneath the crane, invisible from the cab, can be verified through coordinates rather than sight.
The existing anti-collision system is supplemented through integration rather than removed, reducing the burden of adoption.
Automatically accumulated event records become the basis for analyzing hazard points.
Because workers are not identified through video, safety data is secured while reducing concerns about surveillance.
The 3D perception infrastructure provides a foundation for later expansion into integrated yard safety monitoring and autonomous operations.

💡

In a Nutshell

Detecting even the crane blind spots no one can see.

Featured Use Case

Goliath Cranes in Shipbuilding and Offshore Yards

In shipyards, a Goliath crane's cab (on the upper girder) and its actual travel rails (on the ground) are tens of meters apart, limiting the operator's visibility and ability to make safety judgments. LookOut closes this spatial gap with precise coordinate data to ensure safety.

Dynamic Monitoring Tailored to Cranes

Efficient infrastructure: Instead of installing fixed cameras across the entire yard, sensors are installed beneath each crane leg to maximize cost efficiency.
360-degree mobile monitoring: The sensors move with the crane, monitoring the front, rear, left, and right in 360 degrees in real time (focusing detection only on the sections the crane travels through).
Hardware optimization: Wireless data transmission devices and waterproof cables are applied to withstand outdoor environments.

Detection of Unregistered Objects, Without Exception

Wide-ranging detection: Detects not only workers but all moving vehicles on the rails, including bicycles, forklifts, and aerial work platforms.
Flexible response: Detects new materials and obstacles not registered in advance, such as wires and pipes left on the ground, as intrusions without exception.

Excellent Compatibility with Existing Equipment

Cost-saving adoption: Adds LookOut's perception capability without dismantling or replacing existing crane equipment.
Real-time integration: When an object enters a hazard zone, a risk signal is immediately sent to the existing crane anti-collision system and brake interlock circuit.

Scalability Across the Yard

Phased expansion: After integration and performance are validated on a single crane unit, the perception-based system can be expanded across the entire yard.
Smart yard implementation: Maps the positions of transporters, blocks, and workers in real time to understand occupancy, and collects near-miss data to support building a "yard hazard map."

If you need detailed LiDAR crane collision prevention application scenarios tailored to your yard or terminal, please submit a request through the "Request a Custom Consultation" form. Tell us about your crane types, travel sections, and existing anti-collision system setup, and we will propose a sensor placement plan along with an integration approach.

Contents

Use Case

A Goliath Crane Collision Prevention Solution That Detects Even Unregistered Obstacles in Shipyard Yards

Oct 07, 2026

Contents

Who Is This For?

A story for organizations responsible for sites where heavy loads move along fixed paths.

Heavy industry companies operating shipyards and offshore plant yards
Safety management and equipment teams that operate Goliath cranes and gantry cranes on a daily basis
Port cargo-handling companies operating container terminals
Planning and DX departments driving yard automation and smart shipbuilding
Crane manufacturers and maintenance companies, and industrial safety system integrators

They share the same conditions: the objects they handle are heavy, and a single accident can bring the entire yard's operations to a halt.

The Problem on the Ground

While a Goliath crane travels along its rails, four problems overlap.

Safety: The area beneath each crane leg and the travel rails cannot be seen from the operator's cab. The very points where contact occurs are blind spots.
Technology: The objects to be monitored are not uniform. Not only people, bicycles, forklifts, and aerial work platforms, but also wires and pipes left on the rails all count as intrusions.
Environment: The yard is outdoors, fully exposed to rain, sea winds, and seasonal temperature changes.
Decision-making: Situations that came close to contact but stopped in time go unrecorded, leaving no basis for identifying which points are dangerous.

Why Were Existing Methods Insufficient?

The root of the problem lies in what existing measures assume in order to work.

Pre-registration–based anti-collision systems make decisions based on the positions of obstacles entered in advance. Any unregistered object that enters the rails is not detected.
Worker monitoring and hand signals cannot cover the entire length of the rails at all times, and communication is delayed at night and in rain.
Camera-only monitoring relies on 2D images, making it difficult to judge the actual distance and height of objects.
Tags and wearable devices apply only to people. They cannot be used for inanimate obstructions such as wires or pipes.
The absence of records leaves no basis for setting improvement priorities.

In short, there was no solution that simultaneously met all four conditions: working without pre-registration, detecting irregular objects, operating reliably outdoors, and keeping a record.

How LookOut Solves It

LookOut reads the area around the crane in 3D using LiDAR and assesses risk through the following steps.

Detection and classification: People, vehicles, and other objects are detected and distinguished as separate 3D objects. Because LookOut's proprietary AI algorithm detects even irregular objects not included in its training list, intrusions on the rails are recognized regardless of whether they were registered in advance.
Omnidirectional monitoring: The front, rear, left, and right of the area beneath each crane leg are monitored in 360 degrees, capturing zones beyond the cab's line of sight as coordinates.
3D information capture: Distance, direction, height, speed, and movement trajectory are calculated together for each detected object. Contact risks that cannot be judged by horizontal distance alone are calculated with height included.
Tiered hazard zones: Zones divided into danger, caution, and safe are defined with precise coordinates. When work layouts or equipment positions change, only the zones are reconfigured, without moving any structures.
Assessment and integration: When an object enters a hazard zone, the system judges it as a collision risk and sends the result to the existing crane anti-collision system and brake interlock circuit. Real-time video and distance information are also provided through the cab interface.
Sensor fusion: Cameras and LiDAR are calibrated so that 3D perception results are overlaid on the video. Managers can check coordinates and the screen at a glance.
Recording and analysis: Event histories and observation data are stored automatically for later review and analysis.
De-identification: Because people are recognized through 3D coordinates rather than raw video, worker positions are detected without identifying faces.

No pre-registration required, irregular objects included, built for outdoor environments, and a full record. LookOut meets all four conditions above at once.

What Changes

When the premise of assessment changes, the nature of safety management changes.

Because detection does not depend on a registration list, equipment and materials that entered the yard for the first time that day are also detected.
The space beneath the crane, invisible from the cab, can be verified through coordinates rather than sight.
The existing anti-collision system is supplemented through integration rather than removed, reducing the burden of adoption.
Automatically accumulated event records become the basis for analyzing hazard points.
Because workers are not identified through video, safety data is secured while reducing concerns about surveillance.
The 3D perception infrastructure provides a foundation for later expansion into integrated yard safety monitoring and autonomous operations.

💡

In a Nutshell

Detecting even the crane blind spots no one can see.

Featured Use Case

Goliath Cranes in Shipbuilding and Offshore Yards

Dynamic Monitoring Tailored to Cranes

Efficient infrastructure: Instead of installing fixed cameras across the entire yard, sensors are installed beneath each crane leg to maximize cost efficiency.
360-degree mobile monitoring: The sensors move with the crane, monitoring the front, rear, left, and right in 360 degrees in real time (focusing detection only on the sections the crane travels through).
Hardware optimization: Wireless data transmission devices and waterproof cables are applied to withstand outdoor environments.

Detection of Unregistered Objects, Without Exception

Wide-ranging detection: Detects not only workers but all moving vehicles on the rails, including bicycles, forklifts, and aerial work platforms.
Flexible response: Detects new materials and obstacles not registered in advance, such as wires and pipes left on the ground, as intrusions without exception.

Excellent Compatibility with Existing Equipment

Cost-saving adoption: Adds LookOut's perception capability without dismantling or replacing existing crane equipment.
Real-time integration: When an object enters a hazard zone, a risk signal is immediately sent to the existing crane anti-collision system and brake interlock circuit.

Scalability Across the Yard

Phased expansion: After integration and performance are validated on a single crane unit, the perception-based system can be expanded across the entire yard.
Smart yard implementation: Maps the positions of transporters, blocks, and workers in real time to understand occupancy, and collects near-miss data to support building a "yard hazard map."

Contents