Anyone can publish on Medium per our Policies, but we don’t fact-check every story. For more info about the coronavirus, see cdc.gov.

The pandemic shows we need to take back control of our fragile systems

Palantir
Palantir
Mar 25 · 5 min read

Governments and companies need to move past reliance on fragile systems and retake control of our critical structures immediately. Without action, the projections for the impact of this virus could fall short of the true economic, social, and geopolitical consequences.

We are all familiar with the experience of cascading failures. Our flight is delayed because of a brief thunderstorm, but the subsequent slow catering, late baggage loading, crew schedules, and snarled runway traffic exponentially grow that minor disruption into a systematic collapse. COVID-19 is causing whole industries to feel the same pain — once the initial disruption has occurred, the ability of businesses to coordinate across a complex array of functions falls apart. Unfortunately, in a time of crisis, when operating efficiency is most critical, the systems we rely on most perform the worst. The problem is that we have created so many fragile systems that work well when everything is consistent and predictable but spiral into delay and failure when the unexpected occurs.

In particular, the current crisis is revealing the fragility and emergent consequences of large systems that have been decomposed into their constituent parts, incrementally improved over time, but never adequately stitched back together. Without active intervention, error in the system propagates into cascading failures as both the models and the assumptions the models relied upon prove to be false.

Most critically, this is playing out right now in the catastrophic shortage of medical supplies. The more complex the medical supply chain, the less likely that supplies such as masks will reach those most in need — not because they don’t exist, but because decision makers don’t have the accurate view of relative need required to retake control of a system that has run automatically for so long. Other systemic failures are percolating in the background. For example, we are currently experiencing a dramatic decrease in maritime shipping but also, paradoxically, a shortage of available refrigerated containers. This is because equipment movement models assume consistency in vessel movement, and when containers can’t be delivered, they can’t get back to where they need to be. When the foundations for coordination make assumptions that can’t be monitored, you end up with a much more consequential equivalent of a three-hour delay after a 30 minute thunderstorm.

We don’t see the level of resilience we might expect because increased automation necessarily entails simplifying assumptions about how the
world works, glossing over reality. An airline’s revenue management model assumes a static flight schedule and the network model assumes stability in revenue per seat. A global manufacturer assumes stability in demand forecasts and the cost and throughput of manufacturing processes when setting prices or developing a production plan. Increasing sophistication in any one discipline has been coupled with increasing isolation in decision making. Coordination becomes continually more complex even as new local efficiencies are achieved.

The solution to this challenge requires two immediate corrections:

First, we need to create system transparency, not just within discrete, isolated systems, but across them wherever possible. Each government agency and company needs a common operating picture that provides transparency into the state of the system, not just its inputs and outputs. Critically the common operating picture cannot isolate functions or allow for ‘black-box’ operations. Organizations must understand the shifting intersections of human, physical, and economic resources that occur with a rapidly changing environment. It cannot be a purely manual exercise requiring all-nighters and massive human effort to understand the impacts of every new disruption when the disruptions are this rapid and this enduring.

Second, we must take active decision making control back from the automatic allocation of resources based on previously defined and automated models. The current autopilot-style approach limits our control of decisions, both small and large, on a daily basis which becomes incredibly costly in times of disruption. Examine every function that is allocating resources based upon a plan that was developed months ago (when the world looked quite different) and seek to understand how it is likely wrong. The vessel and flight schedule are wrong. The allocation of scarce medical supplies is wrong. The disease screening plan is wrong. The sooner we break from status quo the sooner we can seize the opportunity to redeploy assets and break the negative cycle.

Turning off this autopilot approach and taking control of complex systems is daunting but it is also working inside companies and government that are getting ahead of the virus. Leading shippers, airlines, health care providers and governments are demonstrating that this can be done. In order for us to start, every leader — political, commercial, or otherwise — should make a list of their most critical and scarce resources. Then, we should ask who is deciding where those resources are being allocated, and why? If the answer is based on a previous plan, schedule, or model (or if the answer is “no one”), these leaders must roll up their sleeves, demand the information required to build a common operating picture, and take back control of the decisions around these assets.

We have to assertively wield the systems to make them work for us and not the opposite. This means relearning how to make those decisions we have outsourced to systems. We will certainly get some of them wrong. But successes around the world show that decisions we make with real visibility and live information will be far better than decisions outsourced to an automaton configured for a world we no longer recognize.

Fortunately, this is a problem that can be addressed by smart deployment of technology to augment and support institutional decision makers and the critical analysis they need to carry out to make sound decisions. It is also eminently achievable at pace with existing technology aimed in the right direction. Investing in cross-functional system visibility will provide us a much stronger foundation on which to make critical decisions now and increase the sophistication, and resilience, of our critical infrastructure for the future.

Author

Ted Mabrey, Palantir Global Commercial Business Lead

Palantir Blog

Palantir

Written by

Palantir

Palantir Blog

Palantir Blog

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade