Could Full Context Have Prevented Deepwater Horizon?

We are witnessing what may be one of the worst man-made environmental disasters of all time. As devastating as Chernobyl and even more catastrophic than the Exxon Valdez, the Deepwater Horizon disaster in the Gulf of Mexico increasingly appears to be the result of deliberate decisions to cut cost, and save time.

Other disasters have been easier to classify as accidents. The Northeast Blackout of 2003 cost at least 11 people their lives and, in the first 8 hours, cost $8 billion in economic disruption. The system deficiencies that allowed an outage to escalate into a regional, cascading blackout hid the grid’s deteriorating conditions from the operators charged with maintaining grid stability. To prevent any such future disaster, millions of dollars have been spent upgrading early warning systems, to detect and act upon tiny variances in time series data that indicate worsening system disturbances. The objective is to have sufficient situational intelligence to take informed action quickly.

These electric utilities and balancing authorities had the institutional will to do the right thing. They provided operators the information and the systems they needed to understand, not just what was happening, but the situational context to understand why and what needed to be done quickly to mitigate a problem, and avert or reduce the impact of a looming disaster.

Sub-Millisecond, High Volume Data

Sensor data streams into control rooms in real-time and near-time intervals. Wall-sized and desktop monitors display data and analytics from multiple sources in dashboards and on satellite images and digital maps. Whether in an electric utility, railroad, or oil and gas operations center, the rate and volume of data available can often exceed an operator’s ability to fully understand and quickly act. A high performing person can handle seven plus or minus one or two bits of information in working memory at one time. When operators are fatigued or under stress, working memory drops to the low end of the scale. Confusion ensues. Ability to grasp the complexity of a situation deteriorates.

To compensate for this inherent vulnerability, control room operators require systems that correlate and analyze real-time data from all the relevant sources – sensors, SCADA systems, data historians, GPS devices, weather feeds, enterprise databases and applications – and then display it in an intuitive, geospatial and temporal format. This delivers contextual cues that lead to a faster grasp of evolving circumstances, more informed decisions and quicker action.

Human Psychology and the Unexpected

A landmark psychological study in 1999 asked one group of participants to pass a basketball back and forth as fast as they could. Another group was to count the number of passes. A control group was asked to just watch the passes. In the midst of the experiment, a woman dressed as a gorilla entered the room, walked through the group passing the ball, beat her chest as gorillas sometimes do, and then left. When later asked what they saw, only the control group reported seeing the gorilla. The group that had been assigned to count the number of passes reported never having seen a gorilla. The stress of performing a task in a fast moving situation reduced contextual processing in working memory, as they concentrated on counting the number of passes. The monitoring systems used by operators of critical infrastructure must enable them to see the gorilla.

The Importance of Full Context

In the fable of the blind men and the elephant, six blind men set out to discover what an elephant was. The blind man, who touched the tusk, said the elephant was like a spear. The man who touched the tail said it was like a rope. The man who touched the leg said it was like a tree and so on. They quickly found themselves in total disagreement and confusion. Few stories illustrate the importance of full-context information better than this.

The Illusive Common Operating Picture

True situational awareness poses challenges, but not obstacles. Full context requires correlating data in space and time from dozens of internal and external sources: real-time asset and environmental sensors, weather and event feeds (wind, storm, flood, wildfire, earthquake, sand, cloud cover, lightning, temperature), infrastructure location information, enterprise data, demographics and financial, and other impact data, and geospatial context, including terrain. By displaying that data visually on geospatial images, a fast, intuitive and comprehensive grasp of circumstances is enhanced.

Utilities and ISOs that invested early in giving their operators full-context situational intelligence to prevent disasters have already saved their regions billions in economic disruption. Proactive outage prevention in California during wildfire season keeps the lights on throughout the state. In the coming years, governments and infrastructure companies worldwide will invest 10s of billions more to avert and mitigate crisis situations.

This quarterly column will address the importance of a common operating picture, in full geospatial and temporal context, with intuitive real-time analytics and alerts for infrastructure operations across a variety of sectors.

Mark L. Feldman, PhD, CEO with Margot Rudell, director, Space-Time Insight; e-mail respectively This email address is being protected from spambots. You need JavaScript enabled to view it. and This email address is being protected from spambots. You need JavaScript enabled to view it.