Conversational artificial intelligence has been rapidly growing in popularity lately. Since Facebook opened up its chat API, chatbots have exploded in popularity and are being used for everything from fielding customer services enquiries, to managing transactions. But will these tools gain any traction in maritime?
In the wider B2B space, SaaS scaleups like Drift are leading the charge in conversational marketing, making it possible for any business to build AI-enabled chatbots to field enquiries, qualify leads, and handle customer services. To date though, there has been only a handful of product launches in the maritime sector.
Most container carriers, large logistics operators, or freight forwarders need large customer services teams standing ready to answer the phone and field enquiries. Most of those enquiries usually consist of “where’s my box?”, “why hasn’t my box arrived?” or “when will my box arrive?”. Track and trace is now usually good enough that a customer service agent can answer those basic questions, but fielding the inquiries still takes a lot of staff hours that could be better used elsewhere. Answering simple questions like this is a space where chatbots could make a big difference in operating costs. In 2017, ARL-Shipping launched a last-mile container tracking chatbot that used Facebook’s interface. The bot allows users, mainly shipping agents and warehouse operators, to query the location of containers en route to a warehouse and receive notifications when they are close to arrival.
Ship management is another area where chatbots are beginning to see limited rollouts. METIS Cybertechnology launched its Cyber Personal Assistant a year ago. The chatbot is aimed at fleet management teams and is able to alert users of problems, answer questions, and prepare reports for submission automatically.
Want live bunker prices on your phone? Just add +44 118 310 0466 on Whatsapp as Bunkerex. Send them a message with a port name and their bot will reply with up to date prices in the port. Send a pair of ports, and their bot will reply with all of the prices on the route and recommend the cheapest option taking into account deviation. Finally, send a ship name and a port, and the Bunkerex chatbot will find all of the bunkering ports and prices from the ship’s live position and the port, and again recommend the cheapest.
Chatbots have potential applications in training too. Recognising that Standard Marine Communication Phrases (the IMO’s approved phrases for communication between a ship and shore services like VTS) are poorly used at sea, multiple universities conducted a study to establish the effectiveness of a training chatbot in 2016. In total three chatbot exercises were created for the study, the first offered SMCP training, the second simulated a conversation between VTS and a ship on the approach to Tokyo bay, and the third simulated communication in a medical emergency. 92% of the participants in the study believed that students would have access to the technology required to run the programme and 97% of the participants believed that using the chatbot would enhance the usage of SMCP at sea.
The barriers to creating bots now make it possible to build chat workflows without any code and have a simple bot up and running in a matter of minutes. To test this theory we built our own simple chatbot, it asks users on Facebook messenger what they want from us and points them in the right direction. It took less than an hour to build and costs us nothing to run. You can check it out here if you have a Facebook account.
Though the barriers are coming down and there are a number of novel uses for chatbots cropping up across the industry, we are yet to see any solutions gain real traction. But chat interfaces are relatively young, with the first bots appearing in our lives around a decade ago. Though many see chatbots as another marketing gimmick, as adoption among large B2C companies grows we will likely become much more used to interacting with businesses in this way. While chatbots may not transform maritime services as a whole, the capability to query large amounts of complex data through a chat interface will make them a key tool for delivering those services efficiently in the future.