Why are we still putting up with power outage downtime?
Why are we still putting up with power outage downtime?
One of the biggest causes of downtime and data loss comes not from hackers or other nefarious activities, but from massive power blackouts that are caused by faulty equipment or machine “downtime”. In fact, loss of power leads as the top IT related disaster most organizations face. San Francisco’s massive April power outage, for example, brought high-tech companies throughout the city to a screeching halt, along with the city’s traffic lights, buses, and BART system.
Most businesses haven’t revealed exact losses because many are still negotiating with PG&E Corp., the utility company responsible, but some have claimed losses in the neighborhood of $ 100,000 to $ 300,000, meaning overall losses in the millions of dollars for the city’s busy IT sector.
See Also: eleven-x marches its low-power IoT network across Canada
Power and Infrastructure
San Francisco wasn’t the only city to suffer from a major power outage last April, either. New York also saw widespread blackouts this year, leaving passengers stranded in subway tunnels during the height of the morning commute. This comes after New York had received low grades every year on their Infrastructure Report Card. Still, New York has consistently received higher grades from the organization than the country as a whole, which is worrisome.
Many are asking if this kind of thing will become more common moving forward as the already-strained power infrastructure systems in America’s big cities get older and are overburdened by growing populations. Will power outages be something that companies just have to accept and deal with? I don’t believe so. While these challenges are real and present a serious challenge to power systems, embracing new technologies will allow for utility companies, and their customers, to be better prepared and perhaps avoid power outages altogether.
There is a Crystal Ball
Despite these challenges, power outages potentially not only become rarer but possibly avoided altogether. Using predictive analytics, technology’s crystal ball, both utility companies and the organizations that depend on them will be better prepared to deal with these issues. Predictive analytics will allow personnel to receive early warning notifications when issues appear imminent, allowing them to solve the issue before the power outage event occurs.
Some problems may be identified days or weeks before the power outage would take place. In using predictive analytics to identify these problems, loads could move, and planned outages could occur to minimize damage. Organizations would be able to receive advanced warning of these issues, and take the appropriate steps to mitigate losses during that time frame. Additionally, power companies would be able to identify necessary maintenance costs in advance with predictive analytics. For example, before the system requires maintenance, parts could already be en route.
Maintenance windows would also extend, as predictive systems could monitor equipment conditions to provide more accurate insight into when certain systems need to be replaced. That’s happening now in the service industry. “Using AI and machine learning, we are now able to optimize maintenance schedules, designate high-risk parts for replacement, and ensure technicians have the correct tools and replacement parts so they don’t need to make unnecessary and repeat trips” said Shahar Chen, co-founder of NY-based Aquant.
The real-world benefit of such systems would be difficult to quantify but substantial. The increase in equipment life, improved efficiency, and a boost to productivity would be felt not only for the power company, but for all customers of that company as well.
Predictive Analytics and Acts of God
We may even be able to anticipate the effects of natural disasters on power systems thanks to AI machine learning. Texas A&M University researchers recently developed an intelligence model that can predict the effects of high-speed winds during severe weather so that trees which might come down on power lines in critical areas will be trimmed first. At the moment, this sort of work is done on an arbitrary basis. Predicting an optimal tree trimming schedule is only one of the model’s applications. “Any kind of environmental data that has some relevance to the power system can be fed into this prediction framework, says Dr. Mladen Kezunovic, who developed the system with several grad students.
In areas where recurring severe weather is a factor, Exacter, an IoT company that deals with electrical systems, says its algorithms can identify equipment that’s degraded or showing signs of failure before a major storm strikes. Utility companies then prioritize preventive maintenance to the locations most densely populated and affected. After the storm, they provide health assessments to get electrical systems up and running as soon as possible.
See Also: Nokia helps make Chengdu region into a smart powerhouse
A Win-Win Scenario
Predictive analytics provides a win-win scenario for utilities companies and the organizations that depend on them. By notifying teams of impending risks (potential power outage), they can take the necessary action to mitigate damage and complete repairs. Additionally, companies will have advanced warnings of upcoming downtime and be able to take their own steps. Productivity and efficiency will improve across the board.
What’s really exciting is how AI and machine learning is able to provide such immediate actionable insights. Utilities companies have already started to embrace predictive analytics and will be ramping up their spending in the area in the years to come. According to Navigant Research, utilities will be spending $ 50 billion on grid monitoring equipment by 2023. So while infrastructure and systems do have a necessity for upgrades, our advancements in tech may solve these most issues before another costly, inconvenient outage can occur. At any rate, that’s my prediction.
The post Why are we still putting up with power outage downtime? appeared first on ReadWrite.
(29)
