Algorithms and the new role of the planner in Digital Supply Chain
Machine learning takes supply chain planning to a new level
Algorithms are not new to supply chain management. We have been deploying them for many years in different ways, from simple instructions ('always meet X customer's demand from Y distribution centre') to more sophisticated computations ('meet X customer's demand from the nearest/cheapest/best supplied distribution centre').
Today's algorithms take such computer-assisted analysis and decision-making support to an entirely new level. The scale, speed and technical capabilities of the latest, cloud-based computer technologies is unprecedented. It is now possible for computers to learn from data without being explicitly programmed how to do so. This so-called 'machine-learning' was unimaginable even a few years ago. Similarly, the ability of machines to mimic human neural networks and operate multiple processing layers at any one time – known as 'deep learning' – has rapidly improved classification and prediction accuracy.
What does this all mean for supply chain optimisation? Simply put, it means an infinitely improved ability for operations managers to predict and prescribe everyday eventualities and processes in supply chains.
Does automation lead to doomsday for jobs in supply chain?
The answer is no. This new-found capacity from automated technologies does not make the operations manager redundant. Algorithmic supply chains are different from automated supply chains (although the two often work hand-in-glove). Human decision-making is still required. The difference is the speed and accuracy with which such decisions can be made.
So a new customer order comes in, for example. Should your company accept it? Previously, deciding 'yes' or 'no' would be based on an analysis of limited information and – let's be frank – a sizeable dose of managerial gut instinct. A modern algorithm may well come up with the same conclusion, but in a matter of seconds, and based on a far wider and more up-to-date set of variables.
In short, algorithms and the automated technologies that now often accompany them do not fundamentally alter how synchronised supply chains operate. They simply enable a synchronised system to do what it has now been structured to do – only faster, more efficiently, and with far higher levels of accuracy.
Future planners must embrace the benefits of automated technologies and robotics
As technology advances, so we can expect performance levels to improve. Expect machines on the production line or in logistics networks to produce ever more volumes of real-time data, for instance, as they are individually connected up by WiFi (the 'Internet of Things').
As machines get better at learning from their own experience, expect automation to increase too. Internet retailer Amazon is already experimenting with drone deliveries, for instance. The practice of robots undertaking stock checks in warehouses and shops is one that is spreading fast. Companies like German logistics firm DHL and UK retailer Ocado, meanwhile, are investing millions of pounds in automated systems that retrieve goods or pack items into boxes without human support.
Looking ahead, it's fundamental for supply chain leaders and planners to embrace advances in new data-focused technologies in their processes and infrastructure. It's only inevitable that robotics and automated systems will be integrated into their day-to-day jobs. The competitive obligation to do so is there. Systems change requires rethinking current approaches at a variety of levels. Given this reality, they have to adapt to the accelerated change and start preparing today.
Download our white paper to discover more on how to best manage your supply chain in the digital age.