A Proposed Digital Naturalness Application: Supply Chain Network Resilience
/Our goal with the Digital Naturalness project is to demonstrate the value of an approach to digital technology design leads to increased wellness, beauty, and aliveness in technology users (as well as supports humans to impact nature in a healthy way while continuing to develop more advanced digital tools). To do that we’ve so far outlined a possible set of objective features that can guide design, as well as subjective state-change processes to increase designers and engineers insight and intuition for embedding those features into their concrete projects. Here we suggest a novel product that could demonstrate the usefulness of this approach.
Supply chain risk management is one of the biggest challenges for operations executives. For large companies with complex supply chains, a single crisis can result in hundreds of millions of dollars of lost revenue (O’Connor, 2011). These risks come not just from failures of immediate suppliers, but also failures of suppliers’ suppliers and so on. In fact, supply chain reliability is a function of the resilience of the entire supply chain network. A software tool that could 1. measure and 2. offer suggestions on improving the resilience of a supply chain network could increase the reliability of all participating companies’ supply chains simultaneously. Potential customers for the supply network resilience measurement tool would include companies with complex supply chains and supply chain management software platforms,
So how could we use a digital naturalness approach to building such a tool? Supply chains are complex mutualistic networks. One way of looking to increase their resilience is to ask how other complex mutualistic networks from nature maintain their resilience. Network scientists have looked for useful predictors and drivers of the resilience of complex systems, including ecosystems, for decades (May, 1972) (Gao, 2016).Are any of three technical correspondences to the digital naturalness design principles relevant? As it turns out, coherence is very similar to one of the most frequently cited network characteristics linked to resilience. This characteristic is called “connectance.” There are several ways to define connectance. The simplest is as the fraction realized trophic (i.e. food web) connections over all possible connections. The more of these connections between plants and animals actually exist, the more resilient the ecosystem will be (Dunne et al, 2002) (van Altena et al, 2016) (Thébault, 2010).
So, how could we translate connectance from network ecology and test whether we could build a digital tool that could use it measure and increase the resilience of global supply chains? First we would need to define the relevant entities and relationships and gather the requisite data. For example, the process for determining the resilience of a supply chain for a target industry might follow these steps:
Define Relationships and Gather Data
Define a node in the industry supply chain network as a supplier, client, or transport hub. Define the relationship between these nodes as supplier, client, or transport.
Start with one node. Discover which companies the node sells products to, buys parts from, and ships with (tier 1). For the simplest definition of connectance, only the identification of the connected nodes and their types are required.
Discover which other companies (tier 2) each of the tier one companies sell to and buy from.
Continue with tier 3, etc. as far as data is available.
Repeat steps 1-3 as far as possible for each identified node, in order to discover the total number of nodes in the industry network.
Then, to determine the resilience of the industry supply chain network, calculate its connectance:
Determine Resilience
Connectance = the number of links between companies divided by number of companies squared = L/C2
“Number of links” means how many supplier, client, and shipping relationships are already operating in the network. The connectance measure uses a “directed” graph, meaning that if two companies are both suppliers and customers to each other, that counts as two links.
Calculate the connectance. If there are 40 companies in the analysis and 200 links between them the connectance would be 200/1600 = .125
The higher the connectance score, the more resilient the industry supply chain network.
Increasing Network Resilience
Once resilience has been determined, the software tool could make recommendations to individual companies to build supplier, client, or shipping relationships with nodes in the network with which it does not currently have a relationship. This would increase the connectance score for the entire network, thereby increasing its resilience.
As mentioned, L/C2 is only the simplest of the many related ways to calculate connectance in the field of network ecology. Later, more nuanced iterations of the software could use various weighted analyses by taking into account the unit volumes of the sales or purchases of each company in the network. It could also place companies that have very similar suppliers and customers into sets, then perform the connectance analysis over these sets rather than individual companies. Both of these are inspired directly from alternative ways of calculating a natural ecosystem’s connectance.
Surprisingly, as far as we have seen no supply chain risk management company is using a network topological analysis inspired by resilient ecosystems to measure the resilience of their companies’ supply chains. Using connectance to model resilience could especially aid in reducing long-term risk because it fosters industry wide collaboration and gives companies insight into making the network as a whole as resilient as possible.
This is just one example of applying the digital naturalness design principles to embed some of the deep qualities of beauty, aliveness, and wellness from our interactions with nature into digital systems. As we continue to build more and more complex digital infrastructures that manage more and more of our daily lives and the flow of significant resources, tools like the ones described here, and the process to develop it, will become increasingly useful (Bratton, 2016).
References
Bratton, Benjamin H. The Stack - On Software and Sovereignty. Massachusetts: MIT Press, 2016.
Dunne, Jennifer A., Richard J. Williams, and Neo D. Martinez. “Network Structure and Biodiversity Loss in Food Webs: Robustness Increases with Connectance.” Wiley Online Library. John Wiley & Sons, Ltd, July 10, 2002. https://onlinelibrary.wiley.com/doi/pdf/10.1046/j.1461-0248.2002.00354.x.
Gao, J., Barzel, B. & Barabási, A. “Universal resilience patterns in complex networks.” Nature 530, 307–312. 2016. https://doi.org/10.1038/nature16948
May, RM. “Will a Large Complex System be Stable?” Nature 238:413414. 1927. https://doi.org/10.1038/238413a0
O’Connor, John. “Supply Chain Risk Management at Cisco : Embedding End-to-End Resiliency into the Supply Chain.” 2011.
Thébault, Elisa & Fontaine, Colin. “Stability of Ecological Communities and the Architecture of Mutualistic and Trophic Networks.” Science (New York, N.Y.). 329. 853-6. 2010. 10.1126/science.1188321.
van Altena, Cassandra, Lia Hemerik, and Peter Ruiter. “Food Web Stability and Weighted Connectance: The Complexity-Stability Debate Revisited.” Theoretical Ecology 9, no. 1. January 2016. https://doi.org/10.1007/s12080-015-0291-7.