Based in the maritime hub city of Singapore, Portcast is using machine learning to greatly increase the accuracy of its customers’ cargo demand forecasting.
Founded in October 2017 by CEO Nidhi Gupta and CTO Dr Lingxiao Xia, Portcast has developed an eponymous online platform that combines real-time external data with machine learning (ML). It’s purpose is to bring heightened levels of visibility to its customers’ supply chains.
There are two products that we offer. One is Intelligent Vessel/Container Visibility, where the user is able to receive predictions for when their containers will arrive, and the other one is Cargo Demand Forecasting,” says marketing manager Jia Yung Lee,promising that that the company’s ML-powered predictions have an average accuracy rate of more than 90%.
In terms of forecasting cargo demand, traditional methods typically result in accuracy rates of just 70 to 80%, he notes. Consequently, the greater precision offered by the Portcast platform enables the company’s growing roster of customers, which ranges across the international supply chain from shipping lines and freight forwarders to ports and manufacturers, to enjoy much more effective planning that in turn enables greatly enhanced asset and capacity utilisation.
However, there’s more to it than just that. The Portcast platform also facilitates much longer-range forecasting than is typically the case with traditional methods: around six to eight weeks as opposed to one to two. What’s more, while conventional methods are not only less accurate and more short-term in scope, they are also much more labour-intensive.
The Portcast platform, however, frees a customer’s personnel from such time-consuming activities and in so doing significantly ups productivity, with one shipping line alone saving some 100,000 hours otherwise spent on such tasks in a single year.
Moreover, by using Portcast’s Cargo Demand Forecasting tools, customers are able to gain ready access to such key criteria as load factors, capacity shortfalls and booking trends that in turn can help them reduce costs and improve their yield management. Furthermore, as result of the insights offered by the platform, Portcast customers are also “able to carry out dynamic pricing better” while simultaneously raising their customer service levels.
Meanwhile, when it comes to Intelligent Container/Vessel Visibility, Portcast is likewise able to furnish users with estimated times of arrival (ETAs) and estimated times of departures (ETDs) that are “a lot more accurate than the vessel schedules” presently available to the sector.
Similarly allowing users to switch from reactive to more proactive and leaner inventory and asset planning, this particular tool not only provides users with a visual map interface to keep track of their containers from port of dispatch to port of arrival, but also real-time alerts pertaining to any (potential) hold ups, such as those borne of adverse weather, port closures, strikes or congestion.
And while tracking and tracing might not be exactly new, the accuracy by which the Portcast system can predict ETAs and ETDs is something that Lee sees as very much making it stand out on the market. “A lot of these solutions are focussed on tracking but they don’t really provide predictions,” he says.
Portcast was the winner of the 2019 Captain’s Table Challenge in Hong Kong which is now being launched for 2020.
For other maritime and logistics startup profiles visit the landing page below
