ATSys produced an in-depth engineering investigation for a new Shiploader Collision Avoidance system, and subsequently successfully installed the iSAM LiDAR and GPS system onboard Shiploader 3.
- DBCT shiploaders are manually controlled by an operator on the machine
- Parts of the shiploader structure come into close proximity with ship structures and hatches
- Operator awareness and reaction is required to prevent contact between the shiploader and ship
- Options study developed by ATSys, evaluating similar existing installations and available technology
- iSAM LiDAR and GPS system evaluated as the best fit for Dalrymple Bay Coal Terminal
- Collisions between the shiploader telechute, operator cabin and boom prevented from making contact with the docked vessel
- Improved SCADA and HMI displays, and new 3D view
- Shiploader to ship contact incidents have dropped to zero since installation
Dalrymple Bay Coal Terminal (DBCT) engaged ATSys to undertake a project from 2016 to 2018 to first evaluate options for a new Shiploader Collision Avoidance system, and then install the selected system based on the final recommendations.
The engineering evaluation included research and analysis of existing shiploader collision avoidance systems at eleven reference sites. The report compared the strengths and weaknesses of these systems, taking into account the technology selected, commodity being handled, and environment that the system was subjected to.
The evaluation report analysed the technical issues that each available technology has, and the appropriateness for its selection and use at Dalrymple Bay Coal Terminal. ATSys analysed the proposals of five potential vendors and provided comparisons and recommendations based on these.
LiDAR vs RADAR
All currently available commercial and experimental Shiploader Collision Avoidance Systems use either LiDAR or RADAR as the primary sensing mechanism. Each has advantages and disadvantages that need to be considered, based on environmental and commodity properties.
LiDAR measures the distance to a target using laser light pulses, whilst RADAR comparatively uses radio waves. The wavelength of each is substantially different which affects the performance of the sensors (LiDAR: 900nm to 1600nm, RADAR: 3.9mm).
Coal dust has a particle size of around 1-100 microns and the two diagrams below show the wavelength of each technology against a potential coal dust particle. It can be seen that because the dust particles can be larger than the operating wavelength of a LiDAR system, there may be some reflection off of those particles. Any LiDAR system would require that these particles be filtered out of the point cloud. iSAM were able to demonstrate the LiDAR system operating successfully in a steam environment of similar particle size. This was one contributing factor of that system being selected.
As the wavelength of the RADAR system is significantly larger than most particles, the device can see through these particle clouds more easily. RADAR, however, can encounter issues when trying to measure the distance to a smooth painted metal surface such as ship decks and bridges. As the LiDAR system did not encounter issues measuring the painted metal surface of a ship, this was another contributing factor towards its selection.
Since installation, the iSAM system has successfully demonstrated its ability to filter out the coal dust, and generate a point cloud model of the docked vessel. The scanners were installed in locations that were less prone to dust to assist with this.
Ship Model vs Real Time
In order to calculate the separation distance between a shiploader and ship, the collision avoidance system must convert the measured 3D point cloud into a 3D model. As there is currently no sensor technology capable of scanning the full length of the ship in real time, most vendors have opted to build a 3D model over time as the shiploader moves across the ship, and recall it each time the vessel returns.
The position of the ship is dynamic and moves continually due to swells, as cargo is loaded, and as ballast configuration changes. The ship model, therefore, must be capable of updating in each motion. This is of particular importance at DBCT, as the ships are loaded 4 kilometres off the coast, where seas can cause large movements in the ship position, compared with the shiploader.
ATSys was particularly satisfied with the demonstrated model capabilities of the iSAM system. The iSAM system is made up of two components: real time boom protection, and ship model generation. Each system has separate scanners offering independent protection to a degree. The location of each sensor, including the GPS receivers, which use RTK-GPS to get precise machine position information for geo-referencing of the point cloud data, can be found below.
Location of the sensors on the shiploader boom
Location of the sensors on the shiploader apex
The real time system uses two scanners on the boom and two on the side of the shiploader apex. These scanners protect the top, sides, and underside of the boom in real-time. The ship model system uses two scanners under the boom, with one mounted on a turntable, to build the 3D model over time. The turntable slews 180 degrees in 6 seconds, significantly increasing the scan resolution of the static scanner.
Shiploader Motion Control
Preventing collisions when objects are too close to the shiploader requires the configuration of both motion inhibit and motion slowdown zones. The configuration of these zones is performed via a 3D configuration tool within the iSAM system. The configured zones are overlayed on a 3D CAD model of the machine. Slowdown and inhibit zones are used by both the real time and ship model system.
The iSAM system provides a distance value, and digital slowdown and inhibits signals for each machine movement, based on the zones configured. ATSys integrated these inputs into the existing ControlLogix Shiploader PLC using the EtherNet/IP Socket Interface. The PLC initiates both read and write communications to both the real time and ship model systems.
The health of each of the communications links are monitored independently. If any of the healthy conditions are false then a Communications Fail alarm is generated and the SCA inhibits further machine movement.