Natural disasters can be made more manageable with analytics.
It’s hurricane season. Every day, it seems, yet another weather event threatens the U.S coast. More than 20 major storms so far in 2020, and more coming, many of them hurricanes.
Technology plays a key role in detecting, tracking and warning populations about bad weather. But it also harnessed once disaster strikes.
After Hurricane Maria devastated the island of Puerto Rico, for example, local government agencies used cloud-based applications and business intelligence (BI) to give accurate data to citizens, lower panic, and prevent blatant false reports or rampant scaremongering.
Speed Versus Accuracy
The automated, real-time nature of BI and visualization applications can be invaluable in settings such as sales or customer service. But it can have undesirable consequences in government: an unnecessarily alarmed populace, sensitive information in the wrong hands, and PR disasters based on false reports or misinterpretation of a few data points. Officials become tied up in damage control when their time and effort are urgently needed dealing with the failure of services.
Underlying challenges can include: data appearing automatically with no ability to curate what is displayed; limited ability to separate out who gets access to what data; once entered, no cleansing or verification process; and tools only accessible to BI specialists, data scientists, and power users.
In some cases, disaster remediation needs applications that offer management the power to delay the publishing of a statistic to provide time to remedy a problem or avoid unnecessary alarm. Further, it can sometimes be beneficial to screen data to prevent the publication of exaggerated estimates or blatant falsehoods. Once these are issued, they are immediately viewable by the public and the press. That can set up a snowstorm of protests, attacks and general antagonism that could curtail disaster relief efforts.
In Puerto Rico, Hurricane Maria ripped structures apart and downed power lines throughout the region. Those in charge used BI apps to highlight the current progress of recovery, the volume of customers hooked up to the grid, and the percentage of users with electricity per area. But automatic posting of unverified data posed problems. Estimates of users with power restored were not always accurate. If inflated, some could be given false hope. Accusations of lies, propaganda, and coverups could follow. On the other hand, underestimates could lead to despondency. The press and political opponents would be relentless in showcasing any perceived ineffectiveness in government actions.
To avoid these issues, the local government used a BI app known as Compass by Truenorth. This visualization tool augments visualization and business analytics engines, providing multi-platform application transparency, and automatically controlling information publication.
In the aftermath of a devastating hurricane, this approach helped the government to communicate progress effectively, both internally and externally. It facilitated data consistency and standardization throughout the organization. Instead of publishing a bald report that 0% progress had been made on restoring power lines in a certain province, government representatives received an alert. That gave them time to verify data and receive updated numbers. If false, the true number could be reported. If correct, they had the opportunity to investigate the situation problems and report complete information – in this case, highlighting the valiant efforts being made to restore power despite having to hack through thousands of downed trees and blocked roads that inhibited recovery crew progress.
This example serves to showcase the value of tailoring applications to the real-world environment. Real-time publication can often be invaluable. But delayed reporting is sometimes necessary when disaster strikes and misinformation can lead to lives being lost.
This article was originally published on 10-24-2020