Summary
Key results
Reduced preparation time for cargo handling operations
Eliminated a bottleneck in the unloading process automation
The need to automate key cargo handling processes
In the process of unloading containers in ports, a significant part of operations is carried out automatically by crane control systems. However, under applicable regulations, containers carrying liquids must be handled manually by an operator.
Previously, detecting such containers required continuous monitoring of the process by an employee, who had to react in time and switch the crane control from automatic to manual mode. This manual verification created a substantial operational burden and became a bottleneck in efforts to achieve near-full automation of vessel unloading.
The requirement for visual inspection also limited the maximum utilization of port workstations and made process throughput dependent on the availability and focus of operators. As a result, the pace of vessel handling was directly tied to the human factor.
Target automation would increase port operational efficiency, reduce downtime and shorten vessel waiting times. Faster operations would translate into better utilization of ships and tangible benefits for both the port operator and carriers.
To shorten operation times, reduce the risk of human error and improve the use of port infrastructure, the company decided to implement a solution based on image analysis and automatic detection of containers requiring manual handling. Due to its experience in AI and computer vision projects, Sii Poland was selected as the partner to deliver this initiative.
Cargo recognition system using AI and computer vision
Sii Poland experts developed a dedicated AI solution using deep learning networks to identify cargo types and determine their location based on camera image streams. The system enabled image segmentation and the generation of object contours in real time.
The scope of work included:
- Training a deep neural network to recognize objects relevant to unloading operations
- Developing stable software for continuous 24/7 real-time image analysis from cameras
- Integrating the solution with ABB’s infrastructure controlling the automated unloading process
- Delivering additional software for solution quality control and segmentation
The implemented system reduced the workload of operators, enabling them to oversee a greater number of cargo handling operations simultaneously, while increasing execution quality and automatically recording results in systems supervising the process.
Automation of cargo identification in port logistics
The developed solution eliminated a bottleneck in the vessel unloading automation process – a challenge that had previously not been effectively addressed using standard engineering methods due to varying image conditions and different camera configurations across port workstations. Thanks to Sii’s expertise in image processing and the design of dedicated vision systems, it was possible to deliver a solution tailored to the specific characteristics of the port environment.
In addition, integrating the system with the existing unloading automation infrastructure increased the port’s operational throughput and reduced vessel waiting times, resulting in more efficient utilization of ships.
As a result, the implemented solution significantly reduced the need for manual cargo verification and enhanced the safety of port operations. Process automation lowered the number of errors and shortened the preparation time for cargo handling operations, enabling crews to focus on supervision and technical support.
Key results
- Reduced preparation time for cargo handling operations
- Improved safety and predictability of port processes
- Eliminated a bottleneck in the unloading automation process
- Reduced the workload of port crane operators