Are there risks in using AI to manage risk in port?

AI is poised to revolutionise port risk management but realising its full benefits hinges on addressing several key challenges, panellists at the International Association of Ports and Harbours’ (IAPH) World Ports Conference emphasised. 

Opening a session themed "AI to strategically manage risk," Michael Bergmann, managing director of maritime consultancy BM Bergmann Marine GmbH, said that in today’s data-driven world, "AI is not optional but essential".

Neelima Pamulapati, director of sales at CrimsonLogic, a global technology company specialising in digitalising global trade and part of the PSA Group, introduced CrimsonLogic’s advanced risk management system.

Ms Pamulapati explained that the system combines network, predictive, detection and text analysis models to offer comprehensive 360-degree risk assessments for shipments and seafarers. This system can identify both known and emerging risks, but its success rests on the availability of high-quality data and may require substantial infrastructure changes.

Lars Reger, chief technology officer at NXP Semiconductors, highlighted the role of edge AI.

"Edge AI allows real-time data processing and risk assessment at the source, minimising latency issues," Mr Reger explained.

This scalable approach enables ports to customise AI solutions according to their needs, potentially paving the way for increased robotisation over the next 15-20 years. However, this shift demands careful planning to address energy demands, environmental impacts and the need for workforce retraining.

In the panel discussion that followed, several barriers to realising AI’s full potential surfaced. The panel said data silos, regulatory requirements, high costs, cybersecurity risks and resistance from traditional operations all pose significant challenges. To navigate the challenges, the industry needs standards for interoperability, workforce development programmes, regulatory collaboration and pilot projects, according to the speakers.