Computers Can't Replace Humans For More Difficult Operational Problems Yet: Dr Yvo Saanen

Continuing with their discussions on AI and automation technologies, TLME Future-Proof Technologies Conference moderator Lars Jensen asked Portwise's Dr Yvo Saanen why advanced systems that work so well in simulations are still not able to effectively solve real world problems?

Giving an example of optimizing vessel storage, Dr Saanen said if you expected to achieve the most optimum utlising of stowage space on board a vessel using only an automated system "you will fail".

The AI system will help you quickly identify and utilise 80% of the stowage capacity on board a ship but the rest - more complex - 20% will require the experience and expertise of a vessel planner to best optimise placement of cargo on board.

In this way, the automated system can help free up the vessel planner's time and efforts for 80% of the vessel capacity while allowing him to optimally utilise this saved time to better plan for the balance 20% of space on board the ship.

Giving the example of the Tesla car, Dr Saanen said for 90 to 95% of the time the car manages to drive itself, thus freeing up a lot of time for the passenger. However, for 5 to 10% of the journety the driver's intervention was required "to help the technology".

So, although complete automation is yet some distance away, the state of the technology even as it stands today, does bring some serious cost and efficiency benefits to any kind of transport and logistics operation, Dr Saanen concluded.

BoxBay: Pushing the Envelope to Discover New Frontiers