The World Economic Forum this week published the results of a JUN23 survey carried out among Chief Risk Officers which included a Risk Spotlight on AI technologies. The overwhelming outcome is the request for accelerated AI regulation. So much so, that more than 1000 international experts have signed an open letter calling for a 6-month global pause in AI development while governments across the world catch up and define necessary regulations.
One of the signees, a world leading AI expert and Berkeley professor, Stuart Russel, likens the development of unregulated AI development and use to Chernobyl – a disaster waiting to happen - citing the fabrication of fake news, racist output, or providing instructions on building weapons to just anyone, as just some of the risks involved through the availability of unregulated AI systems. For Joe/Jane Bloggs on the street, with a limited understanding of the true application of AI, the biggest fear is that it will lead to them losing their job.
What about AI in air cargo?
The past 5 years in particular, have seen our industry finally begin to embrace digital enhancements and the understanding of where AI can drive more efficient, safer, and more sustainable operations. A growing number of SaaS providers are developing and augmenting digital air cargo solutions, increasingly introducing AI-enhanced modules, whether these are for load-planning, pricing, route optimization, or tracking, to name but a few. AI is able to analyze a large variety of information such as historical data, weather conditions, traffic patterns, market dynamics, weight distribution, space utilization, safety regulations, available resources, etc., within seconds and enable decisions that lead to better use of available capacities, greater fuel efficiency, shorter handling times, and lower operating costs, for example.
The benefits to our industry are many – not just in sales, handling, ground and flight operations, but also with regard to aircraft maintenance (AI-powered predictive maintenance can monitor the health of aircraft components in real-time, identifying potential issues before they lead to failures, for example – thus minimizing downtimes, reducing maintenance costs, not to mention improving overall safety), regulations compliance, risk assessments, fraud detection and security, and above all, customer-related services such as tracking, monitoring, and customer inquiries concerning status updates or general information.
Where there are positives, there could also be negatives
Without a doubt, and as proven by a growing number of airlines, forwarders, and ground handlers adopting AI-supported software as it begins to advance in an industry that has often been stamped as one that is “late to the party” when it comes to digitalization, AI offers numerous benefits. Yet, it would be naïve to ignore the potential dangers and challenges that AI in air cargo could pose. These could include, safety concerns, for example. AI systems are not infallible and can make mistakes. Faulty, biased, or inaccurate results produced by AI algorithms, could lead to incorrect decisions on critical processes, which could have a knock-on effect on the entire supply chain. Cybersecurity risks, too, could become problematic: the more AI becomes integrated into air cargo operations, the more vulnerable the industry may be to cyberattacks. In the worst case, hackers might gain unauthorized access to cargo data, flight information, or even take control of cargo-carrying aircraft. Similarly, data privacy could be at risk, given that AI relies on vast amounts of data for training and decision-making. This sensitive cargo information could fall into the wrong hands, for example. The same risk applies during the exchange of information among the many involved air cargo stakeholders: airlines, airports, forwarders, ground handlers, logistics companies, customs, or other partners. Ensuring seamless integration and interoperability of AI systems across these entities can be challenging. Also, should AI eventually lead to the redundancy of people in the process, another AI risk could involve the lack of ‘Human Oversight’: If we depend on AI always providing the perfect answer, we might fail to spot critical issues which could lead to dangerous results.
We are still a long way away from any ethical dilemmas such as self-driving cars already face, today. However, a future scenario could raise ethical questions about responsibility and accountability: If an AI-powered cargo aircraft is involved in an accident, who is responsible—the manufacturer, the operator, or the AI itself?
To address these dangers, it is essential to have robust safety protocols, thorough testing, regular audits, and transparent communication with stakeholders.
Cargo is a people business – will it still be so in future?
Job displacement is likely for those jobs that lend themselves to automation and robotics – these could be in ground operations, cargo handling, data entry personnel, and customer service representatives, to an extent. On the other hand, our industry is often struggling to find staff, so AI is more likely to be a necessary and welcome bridge to the increasing staff gap. ‘Dehumanization’ of the air cargo industry through AI-powered chatbots and automated systems taking over human customer service tasks, could lead to a less personalized and less empathetic customer experience in theory. Yet, looking at ChatGPT and the ever-evolving language systems, it is often hard to tell that a human isn’t behind the answer appearing on the screen. That said, reduced human manpower can lead to loss of expertise in the industry: With AI handling complex tasks such as route optimization and demand forecasting, possibly resulting in less human interaction in these areas, this could potentially lead to a loss of specialized knowledge and skills. And, as stated before, if human operators rely too heavily on AI telling the ultimate truth, they may lose the ability to spot and react to critical circumstances and unforeseen situations.
Human-centric approach to AI in aviation
The global call by Stuart Russel and his peers to pause AI development for 6 months, to give governments a chance to come up with regulations, is – in my view – illusionary. Personal interest will see AI systems continuing to develop, regardless, and not all governments are going to decide on AI regulations at the same pace, scope, or share the same interest everywhere. Indeed, whilst many often have limited true understanding of the subject matter, others will have their very own agenda for AI. Current geopolitical events are proof enough that rules are not kept to if they do not serve a country’s strategy.
Coming back to air cargo – or aviation in general, the good news is that EASA, for example, has been working on regulating AI in aviation since 2018, and recently published its “Artificial Intelligence Roadmap 2.0 - Human-centric approach to AI in aviation,” which outlines the EU AI strategy and its application in aviation. It is a dynamic paper that is updated as the AI situation unfolds – and this is the logical and healthy way forward when it comes to regulating.
On the cargo side, the various SaaS providers are clearly focused on Technology for Good, and are bringing the positives to the industry and to the humans working in it. To summarize, these are:
- Improved Safety - AI augments human decision-making by providing valuable insights and predictive analytics, leading to safer operations.
- Increased Efficiency: Through the automation of repetitive and mundane tasks, AI frees up human resources to focus on more strategic and creative aspects of their roles.
- Enhanced Customer Experience: AI-powered systems provide quick and accurate responses to customer inquiries, leading to improved customer satisfaction.
- Better Resource Allocation: AI's ability to analyze data can optimize resource allocation, leading to cost savings and more effective utilization of assets.
As for those out there worrying that AI is dehumanizing our industry entirely, I believe the record numbers of human attendees at the various cargo conferences over the past year, is testament to the fact that air cargo is and will always remain a people industry. One that is increasingly supported by its ‘Friend’, AI.
We welcome and publish comments from all authenticated users.