The Shipping Industry Is Lagging Behind on Embracing AI
October 24, 2022
Artificial intelligence can provide forwarders with an opportunity to enhance their decision-making capacity. Data transmitted by commercial vessels, for example, can be used for predictive guidance and performance monitoring, enabling businesses to forecast greenhouse gas emissions. This is a way to lower emission levels and ensure carbon transparency. Other implementations include automated processes, improved scheduling, real-time analytics, and so on.
To leverage AI, however, international freight forwarders should take multiple factors into consideration, including the legal framework guiding artificial intelligence. One particular challenge to implementation, especially for shippers operating in both EU member states and the UK is the difference between proposed EU and UK regulations. Other obstacles to the adoption of AI include challenges in data collection, the talent shortage, and lack of management buy-in.
The EU and UK Approaches
In the United Kingdom, the focus of the proposed regulations is on stimulating innovation. The proposals outline 6 principles to ensure fairness, transparency, and safety. Regulators will be given the leeway to develop more specific interpretations that align with different industries. In comparison to this “light” approach to AI governance, the proposed EU AI Act sets out prescribed obligations and requirements for users and developers, guiding some specific uses. The new AI Liability Directive also includes guidelines on claiming compensation in case AI systems or software cause damage. The AI Act complements the directive by specifying high-risk systems, examples being the operation and management of critical infrastructure. Such systems will be governed by more detailed requirements and obligations with regard to implementation.
Other Obstacles to Adoption
In addition to regulatory discrepancies, there are barriers to adoption such as huge volumes of data and poor quality of information. The lack of reliable data at every stage of the supply chain is one obstacle to widespread adoption. Any technology is as good as the data it can access. Some businesses collect huge volumes of information that can be full of redundancies and inconsistencies, leading to data decay. Streamlining collection processes through labelling, data cleansing, and warehousing can help improve accuracy.
The AI talent shortage is also a pressing problem for many organizations. The massive growth in demand means that talent with the right skillset would ask for top-level positions and 6-digit salaries. Not only this but giants that have the knowledge and resources to invest in AI are seen as offering more room for career growth by those who have the training and skills. Think of Baidu, Facebook, and Google.
Thirdly, the lack of understanding of AI among senior and mid-level management can hinder adoption in many businesses. At present, the key challenge for managers is to learn how to leverage AI in specific contexts and scenarios and how to address the risks and limitations of the technology to make ethical and reliable decisions. Data science, statistical thinking, and risk literacy are some of the areas where portfolio, program, and project managers need more training.
