Resilience engineering and what it means for the future of manufacturing systems
The Covid-19 pandemic revealed a fundamental weakness of modern supply chains and production systems: a persistent trade-off between efficiency and resilience.
Over the past few decades the lean production paradigm, along with the closely related just-in-time production, has dominated modern supply chain management and manufacturing systems. Broadly speaking, the core idea behind lean and just-in-time production is to enhance the efficiency of production systems by minimizing buffer capacities and inventories, thereby reducing production costs.
As the pandemic showed, however, the increase in efficiency comes with a trade-off at the expense of a system's resilience. When governments shut down the world to combat the spread of runny noses, supply chains collapsed. Manufacturing systems based on the lean production paradigm were particularly vulnerable as their reduced inventories and lack of buffers left them unable to adapt to sudden disruptions.
In essence, resilience refers to the capability of a system to absorb disruptions, avoid system failure and recover to its original state within an acceptable timeframe. Within the context of resilience engineering there are two high-level strategies that can improve a system's resilience, which include redundancies and flexibility.
Redundancy and flexibility
Redundancy means that a manufacturing system has backup resources, which can be used when primary resources are not available. Redundancy can be achieved by different means. For instance, under normal conditions a system might operate at 60-70 % capacity, allowing it to ramp up the production process by 30-40 percentage points in case other machines are offline. Redundancy can also mean maintaining larger inventory stocks, which can absorb disruptions in supply chains and material flows.
The second strategy for designing a resilient system is to increase the flexibility of a manufacturing system. This can be achieved by interconnecting and integrating various production lines and components, creating a system in which resources can be reallocated and scaled dynamically as needed.
The rise of cyber physical systems and such technologies as the Internet of Things (IoT), cloud computing, big data, artificial intelligence etc. provide new solutions and tools to enhance the flexibility and resilience of modern manufacturing.
Digital technologies allow manufacturers not only to integrate the machines at the shopfloor level but also the entire supply network. In such an integrated system the information does not follow a unidirectional flow, but rather flows in all directions, forming a feedback loop and integrating the entire supply chain into a production ecosystem.
Challenges for the future
Despite the availability of resilience strategies and mechanisms that mitigate the impact of disruptions, implementing resilient manufacturing systems is easier said than done. Building resilience involves a range of trade-offs and requires specific organizational and technical conditions.
First of all, resilience is not the only target variable of a production system as other variables such as efficiency, quality, costs, regulatory compliance, standards, sustainability etc. need to be addressed as well. Furthermore, resilient systems require a bunch of organizational preconditions and enablers, including top-level management commitment, financial investment resources, a reporting culture, technical expertise, transparency, awareness, access to reliable data across the entire manufacturing ecosystem. Moreover, resilience parameters and other target variables must be measurable and clearly defined.
Since modern supply networks and production systems follow the lean paradigm, which emphasizes efficiency and cost reduction by decreasing buffers and cutting inventories, probably the most important question is how supply management systems can be reorganized and how trade-offs are rebalanced? Are resilience and the lean production method inherently incompatible or is there a way to reconcile the two approaches?