Will bunker buying be revolutionised by AI?
Todays bunker intel newsletter to customers considered this question.......get in touch to be added to our distribution list.
This is a simple question with a very complex answer. In essence it is yes and no.
AI will come to the fore to analyse huge amounts of data and to then suggest best strategy. We see this already happening with applications such as Stormgeo's bunker planner. The issue with AI comes with the following- you only get out what you put in! If you put in inaccurate data guess what, the AI algorithms will throw out the wrong suggestions. Some providers understand this, Stormgeo for example, have incorporated NSI's bunker intel data into their platform ensuring in the top 60 ports by volume they have accurate data.
Other companies are basing AI systems on pricing algorithms rather than prices got by individuals who know the market. It is the equivalent of building the Burj?Khalifa directly on a pile of sand.
Where AI will not overtake human interaction is the actual getting of prices. Whether we like it or not the bunker market is determined by human relations. There can be a price difference of up to $10pmt or even a no quote dependent upon who you are and which company you work for. AI will not throw up the suggestion of doing someone a favour when they are long on product or to substitute a vessel due to a cancellation to avoid a big charge. These are examples of the intricacies of the market.
So where does this leave us? I would argue AI leaves the market with risk. If we discount the human factor too much we may become over reliant on tech and not take the time and effort to train staff properly. This involves them learning the market but also visiting clients, suppliers, conferences, bunker hubs to get a real life feel of the market. Most importantly it builds relationships. If then in conjunction to 'dumbing down' AI decision making is based on faulty data sets then we are really exposing both the buyer and supplier to risk.
Another example is with claims. How many claims are solved through a frank discussion and a little give and take. These nuances are not found in simple density discrepancy analysis for example. A testament to how the market works is the high percentage of claims that are resolved before going to court. Again experience and knowing when and how to push is key. AI may well call for specialisation and? I predict this will come at the expense of expertise if we are not careful. Just as Henry Ford all those years ago rationalised motor production to create a plethora of boring jobs that could be done by less skilled labour AI may well bring a modern day equivalent. Rationalisation is fine when what you want is to create a uniform product like Ford's. The added efficiencies mean you can make more, of a better quality and at a faster speed. In a market where there are a million different inputs I would argue rationalisation without experience and intelligence will only result in a mess and higher prices.
Before sending this I shared with our team, and Mike in our Dubai office had a slightly opposing view……
AI can drive a car, fly a plane and pilot a ship (although an AI vacuum cleaner still cannot navigate around a belligerent sleeping cat)…. It is inevitable that AI will become the driving force in shipping decision making matrix in years to come, but we are a way off from that point.
We were impressed around 25 years ago when distance tables were digitalised and further iterations built on custom distances and then route planning and then route planning with ECA’s and adding more and more functionality as the years have progressed. But ultimately the cost of developing AI for a role or task has to be less than the human time performing the role and perform it to a higher and more consistent standard. This is where I see the challenge in the complex shipping markets, until AI develops in other sectors of the market, cargo blending, cargo pricing, ?planning, chartering, long term weather forecasting, freight, demurrage etc then each sector is held back as human interaction is needed to adjust for these inputs. We all know bunkers would be a much easier job for everyone if the charterer would just tell us where the ship is going! So until AI fixes that part of the chain, as humans I think we still have value in this market. (Sorry Charterers I’m not picking on you, it could also be weather, port scheduling, next voyage, etc etc – but you are an easy target on a Friday)
领英推荐
I firmly believe that we are at the beginning of the AI revolution for shipping and there are rewards today for those that embrace the best accessible technology alongside highly skilled human interface. I think we have a market leading, affordable, simple interface, proposition for our customers and we will stay at the forefront of independent development.
But ultimately…. At some point (20-30 years time), I'm sorry to say, I think the vast majority of us reading and writing this will be replaced.
________________________________________________________________
As this is such an emotive subject am very happy to hear your views on how you see the market developing.
Paul Hardy
Business Development
Nautical Supply International Limited
Juniper Court, Boxwell Road, Berkhamsted, Herts, HP4 3ET
T: ?+44 (0)1442 291 430??????????????? ?Skype:?paul12345_1
M:?+44 (0)7807 722 453??????????????? ?Ice:???? ?phardy4
#bunkers #ai #shipping #freight #maritime #marinefuels #chartering #tankers #drybulk
Bunker Broker
9 个月Definitely Brokers Owners Suppliers need to exchange infos communicate together. This can be done on the phone or over a beer or a cup of coffee. The brokers role is to select info and share it with the customers. But yes nowadays to be informed is to know the market or understand it ? It s likely to be both.