As Artificial Intelligence makes inroads into contract logistics, there is a growing need for a reality check on how well prepared the industry is to adopt the technology, and what the next steps are.
Kris Kosmala, Quintiq, Vice President Asia Pacific, talks through the challenges and opportunities.
In what areas of contract logistics will AI have the biggest impact?
First and foremost, keep in mind AI has many disciplines, some of which may be of use in the contract logistics sector. The one area that is always mentioned regarding analytics, which is most useful in logistics, is Machine Learning.
Machine Learning has had an impact on the entire chain of cargo flow planning and execution. It impacts Demand Understanding and Demand Acceptance – i.e. the machine indicates which customers are good now and which will be good later; which commodities will generate how much in profits; and what prices to charge that maximise profit (under contract and spot buy conditions).
What are the major barriers to the successful introduction of AI to the contract logistics sector?
Fear of unknown, lack of imagination, lack of internal technology and technology vendors’ inability to explain the value of AI versus the old trusty Transport Management System with very basic rule engines. Another major problem is that shoddy data is stored in ERP systems, and there is a train of thought in IT departments that cleaning all this old data for the fancy AI is not worth the time and effort.
Is there a need for the industry to establish standards for AI technology in order for greater connectivity between supply chain partners?
There are no standards for AI (it is maths after all, and very ‘standard’ in the way it behaves), but standards are needed to exchange the data within the provider’s internal systems and those of their partners.
We already have many standards for this from the International Port Community Systems Association (IPCSA) and the International Maritime Organization (IMO). However, we need a few updates and, most importantly, a deeper definition.
For example, if you consider the Real Time of Arrival (RTA) for ships and containers – is the RTA set when the container is unloading, or loaded on the prime mover, or stacked in the yard, or loading on the truck-trailer?
To start counting dwell time (and the cost that will be charged to the shipper), logistics providers must have this agreed with the carrier, terminal operator, transporter and importer, and pass RTA without any additional translations along the information cargo flow chain.
Note the comparison with EDI, which forked out into adaptable programs such as ANSI, EDIFACT, TRADACOMS and ebXML? Blockchain faces exactly the same with the same ‘forks’ in the road ahead. Acceptance of standards starts with the decision to computerise. Many freight forwarders working with Logistics Services Providers are behind the ‘8’ ball on computerising to the point of following standards for their internal documents.
However, unless each player on each trade lane for each type of cargo commodity is on the same page electronically, we are going to have issues, no matter what the blockchain pushers tell the audience of logistics people and the banking, insurance and customs players behind them.