Improved productivity in container handling operations

By Dr. Prasanna Lokuge
Chief Manager,
Information Systems Division
Sri Lanka Ports Authority

International trade is governed by shipping as the most important mode of transportation. Both the world’s shipping fleet and seaborne trade has experienced continuous growth over the last few decades.

Worldwide container trade is growing rapidly and it is anticipated that the growth in containerized cargo trade will continue as more and more cargo are transferred in containers.

Ports around the world are icreasing their capacity to handle more container ships, where information technology will play an important role to speed up the turnaround.

By 2010, it is expected that 90 percent of all liner freight will be shipped in containers. Every port is expected to increase their capacities in handling cargo operations in a more efficient manner to gain a competitive advantage in this highly competitive business. Services provided in a container terminal must be improved to assure high customer satisfaction.

Terminal operators are trying to attract more vessels by assuring minimum operations time at the berths, automating equipment handling services, furnishing electronic means of data transfers, minimizing the waiting time for berths and assigning priorities in vessel berthing.

At the same time, they need to reduce the cost of operations, assure the utilization of resources such as Gantry cranes and transtrainers, and prime movers in the terminal.

The use of conventional software techniques to solve such problems would incur high implementation costs and it is difficult to do so since intelligence and human intervention is required in managing the dynamic behaviour of such systems. Continuous monitoring of the ongoing vessel operations at berth is very important, as there may be many unforeseen events that could affect the original plan.

In a generic berthing system, it is required that the most suitable berth be found for the calling vessels. For the allocation of berths it is minimally required to consider: the drafts of the vessels, the crane outreach requirements of the vessel, expected vessel productivity of berths, waiting time of the calling vessel, skills and experience of the staff, the length of the vessels, the type of cargo in the vessels, the priority required in vessel operations, the expected completion time of the on going vessel operations, the expected time taken for loading/unloading operations.

The complexity of the problem is enormous as there are many internal and external factors which govern the decision making process. There are various inter-related decisions made during vessel operations and in berth selection processes, which are regarded as extremely complex due to the dynamic nature of the application.

Therefore, it is essential to introduce intelligent adaptive systems to obtain higher productivity at the container terminals. Several important operational aspects have been addressed in previous studies carried out in this field. The lack of intelligence and interactive learning capabilities in these works hinder their complex dynamic vessel operations in a Port.

Information systems have been undergoing transformations ever since businesses started to use computers for business systems. The earliest systems were data processing systems, which did not place much reliance on information processing.

The developments in software, hardware and information systems architectures brought forth greater revolution in the business world.

Centralized computing, distributed computing, network computing and deployment of Internet application have all contributed to the development and enhancement of business systems. Transaction processing systems, Management information system, Decisions support systems, Expert systems and EDI have added to the value of a business and have been catalysts for the growth of a business.

These systems have increasingly become software intensive and in the latest revolution artificial intelligence has been built into software for autonomous decision making. The future of artificial intelligence is the building of intelligent software agents, which simulates human reasoning and thinking process.

What are Artificial Intelligent Software Agents?
An intelligent agent is a knowledge-based system that perceives its environmental behavior to interpret perceptions, draws inferences, solves problems, determines actions and acts upon that environment to realize a set of goals.

One goal of the Artificial Intelligence (AI) community is to build computer programs that can show autonomous behaviour in real life, that can independently make good decisions about what action to perform and how to execute these actions. In other words, the need to create rational agents that meet the requirement of autonomy has recently become a more prominent research goal of Artificial Intelligence.

The resultant systems are often called "Intelligent systems" because they are capable of directing their behaviour independently in a dynamic environment.

For example, artificial intelligent agents should be able to observe any deviation of container loading/unloading speed of a vessel and suggest suitable alternative methods to improve the productivity of the vessel at the berth, or predict the optimized vessel-berthing schedule when there are variations to the original vessel schedule set by the terminals.

Human practical reasoning appears to consist of at least two distinct activities. The first involves deciding what state of affairs we want to achieve.

The second involves deciding how we want to achieve them. The decision about what state of affairs to achieve is known as deliberation. The process of how to achieve these states of affairs is known as means-end reasoning. Artificial Intelligent agents proposed for the container operation simulate the human practical reasoning and human brain in making decisions.

Importance of using Artificial Intelligence Systems in container ops:

• Personalized Services. The past experience of handling similar vessels is useful to improve the operations of vessels in the berth. Further the characteristics of the vessels in a particular service, past performance and drawbacks are considered in the software agents. Present systems do not use artificial intelligence to provide valuable information to terminal operators with regard to this issue.

• Less human intervention in decision-making. Terminal operators report the expected time of berth, the completion time and the designated berth, to shipping lines prior to their arrival. If for example, a crane breaks down in the berth, they need to find whether the designated berth is still the optimal solution, or whether any other berth can be assigned to achieve the agreed performance for the vessel. In the present systems, human intervention is very often required to find an alternative option for a situation such as this.

*Autonomous behaviour in selecting alternative berths. Once a berth is allocated to an incoming vessel, is it possible for terminal operators to find alternative berths for it in case the present berth becomes unavailable? Present systems do not provide the facilities for selecting alternative options without human intervention in the terminal.

* Monitoring facilities of the optimized solution while on the job. It is required to find optimal solution during the run time of the system. While most systems can find the solution at the designated time, the lack of flexibility for interactive learning in the present systems… generates poor results in the vessel operation systems. For example, systems should be able to monitor the progress of the vessel operations at the berths from time to time and prompt any deviations or alternatives in maintaining the targets set at the beginning.

*Present systems lack the flexibility to deal with uncertainty in the environment.

Most of the data and information used in real shipping applications are vague or incomplete. For example, weather conditions, skills of the people, performance and the ability of the cranes, trucks, stacking/un-stacking time durations. The ability to deal with. environmental uncertainty is very poor in the present applications.

*Autonomy in selecting the best possible resource combination. Intelligent agents could dynamically select the best possible combination of people, number of cranes and trucks required to optimize berth utilization and to assure early completion of the vessel operations.

* Improved Customer relationship management. Intelligent agents help to predict any delays or other issues with regard to vessel operations in advance. This helps to update the respective customers on a regular basis. The negative and positive impacts of: the time taken for lashing/unlashing operations, crane productivity, loading/unloading issues, time taken for stacking and un-stacking operations, synchronizing of yard operations with vessel operations, performance level of the people and time taken for berthing and sailing can be dynamically monitored by the intelligent agents

Professionalism and knowledge users
Implementation of new lCT systems face enormous challenges in any industry and there's no exception in the shipping industry. Although the IT personnel recognize the need for new systems and subsequent changes in the business processes, many 'users are reluctant to accept the modern IT methods to replace the existing legacy systems. One of the major challenges that we face today is the change of attitudes of users and the inertia shown in becoming knowledge users. Our vision is to produce mature users through seamless transition to artificial intelligence systems assuring efficient services to port users and stakeholders. We hope to work with operational change management teams to address the structural and/or bureaucratic barriers in order to streamline lCT related processes and implement artificial intelligent systems for improved port productivity and customer relationships. As software engineers we are happy to take this challenge without any delay for the betterment of the shipping industry and Sri Lanka.