Thanks to modern communications, news about disruption, drama and death is moved around the world in seconds. But bad things have always happened, and although it often took months to share the news elsewhere, it didn’t mean that the world was less risky than now, in fact in many cases, it’s the opposite.
The fundamentals used to price and manage risk were based on ‘what needs to happen’, ‘where will it take place’ and ‘who is going to do it?’ Most of these questions coalesced around the planning for long journeys, often by sea, and the trading of manufactured goods. Investors were always seeking as much information as they could get to help them quantify the levels of risk involved. They then tried to mitigate the risk by working with people who had done similar things before and had survived, or had not gone broke in the process.
In other words, risk was determined by experience and knowledge. The information helping to create the required knowledge generally took a long time arrive and the necessary experience often took a lifetime to accumulate.
These days, knowledge and experience are still critical in understanding risk, it’s just that access to data, transforming it into information and using it to comprehend risk, has been compressed into a real time stream actionable intelligence.
Companies operating or supporting global supply chains, now have to navigate a myriad of changes and disruptions that seem to occur on a daily basis. Because news travels fast and bad news even faster (as they say), logistics operators are constantly managing alerts, alarms, notifications, update requests and customer delivery demands each day. They are doing this against a background of unplanned system failures in their own or partner operations, inaccurate data, key performance metrics and increasingly, geopolitical events that erupt without warning. In this context, it is impressive that supply chains continue to function at all.
Despite the pejorative inference above, information systems are critical in helping supply chain managers and logistics operators do their jobs. The very best systems are not only able to support and qualify the constant flow of data into the supply chain, but to process it into actionable intelligence. This can then be used to inform decision making in real time. In some cases, instances of machine learning can automatically adjust operational systems without bothering the human, subject to the rules inherent in the machine learning algorithm.
The use of intelligent systems to provide decision support in order to augment supply chain operations management will only get better. But to do so, it requires a few key elements. Critically, any data either generated or captured by activities across the supply chain, must be accurate and have some context. The context might include when it was generated, where, by what system and in response to what stimulus?
Thanks to the ubiquity and continuous manufacturing breakthroughs related to mobile phones, the price of sensors has collapsed to almost nothing. This means that it is possible to add sensors to almost every item moving through the chain. As 5G technology is rolled out across the world, it is providing the communications platform that can support the billions of new sensors ‘beeping’ and ‘burping’ location and status updates as required. The only drawback to this revolution, is that a large number of the operational systems currently in use, were never designed to deal the volumes of data now coming through the pipes.
This could also be quantified as an operational risk, in that the inability to receive critical data about a problem, removes the ability to avoid it and increases the cost of resolving what happens after it occurs?
If well designed and implemented operational systems are used to manage a global supply chain, by definition, that supply chain is a less risky proposition. The flow of products and information should be consistent with operational metrics and any issues or problems will be detected at an early stage, enabling swift correction or alternate solutions. The operators should have more time to consider how they manage their resources to either improve performance, or increase capacity. They should also be able to respond and manage any significant disruptions due to unplanned geopolitical events such as war or natural disasters (or new pandemics?).
The abundance of data and information that will accumulate within such an operation should be an ideal reference for pricing risk for insurance purposes. Indeed there are a few companies emerging that have developed artificial intelligence models that will price risk in real-time for various aspects of supply chain and logistics operations. e.g. individual transport legs, manufacturing machine performance, quality assurance related to product sourcing, etc.
Advanced vision systems are already augmenting the pick, pack and fulfilment operations of Amazon, identifying damaged or miss packed goods before they are shipped to the customer. This has reduced claims, cut expensive return and reshipping costs and helped the work with shippers to improve product quality. It would be surprising if this use of intelligent systems is not replicated in numerous other aspects of fulfilment operations.
Therefore, as information systems platforms capable of absorbing the huge volumes of sensor data generated by supply chain operations proliferate, it is an optimistic hope that things will get better, faster, and more resilient.
In short – be less risky!
Author: Ken Lyon
Source: Ti/Foundation for Future Supply Chain
