How AI/ML Help Secure the US Power Grid Infrastructure

Print Friendly, PDF & Email

The power grid of the United States is one of the most complex and technical systems in operation around the world. In order to deliver consistent electricity to the entire country, a number of regional transmission organizations (RTOs) must interact and manage resources.

Like with any wide-scale network-dependent system, the electric grid is vulnerable to cyberattacks from outsiders. Hackers may be looking to cause disruptions in service or may have a larger goal of affecting the supply chain of energy resources.

The U.S. government and the electric and gas companies are now moving into a more technology-focused future where new sciences like artificial intelligence and machine learning can be leveraged to help secure the power grid, its infrastructure, and customers nationwide.

Managing the Grid Historically

The RTO system was first instituted by the U.S. government in 1999. In essence, it operates on standard supply and demand model. Utility companies deliver their supply of electricity to the RTO in their region, where it is grouped together and priced accordingly.

Before electricity is distributed across the regional grids, a computerized marketing algorithm is run to determine different price points and area needs. The individual blocks of electricity are sold to customers until the demand is met. In fact, the energy that reaches your home or office may actually originate from a plant located far away from you.

Up until the last few years, the computerized system for managing the power grid was heavily dependent on manual configuration and maintenance. In addition, there were a number of inefficiencies that could lead to shortages or congestion in electricity service.

Worst of all was the risk that hackers could infiltrate the grid network and potentially wreak havoc on the entire system.

Moving Towards a Smart Future

There are two significant reasons why the U.S. government and utility companies are aiming to migrate to a smarter grid. One is to use big data to deliver electricity in a more efficient manner. The other is to keep the grid’s infrastructure more secure and reliable.

Saving with smart meters: Many utility companies provide customers with smart meters to install at their homes or offices. These devices are part of the internet of things (IoT) movement, which aims to connect more hardware and equipment to the world wide web.

Smart meters have real-time sensors that track electricity usage and climate conditions. All of this information is fed back to the power provider and is used to optimize pricing and minimize congestion on the grid.

This is critical because current studies estimate that 61% to 86% of energy on the U.S. grid goes to waste.

For the first time in history, data related to the power grid is being communicated in a two-way manner. Utility companies gather as much information as possible to make more informed decisions and provide better service to their customers.

Modern Threat Prevention: But of course, efficient delivery and lower prices do not help anyone if the utility companies and RTOs cannot defend themselves against cyberattacks aiming to take the system down or destroy it. If even one piece of infrastructure on the central grid becomes compromised, the rest of the system is at risk and so are all citizens.

With increasing confidence, the United States is turning to artificial intelligence and machine learning to supplement how humans manage the grid. The idea is to teach computer systems what normal activity looks like so they can easily detect anomalies and raise alarms when necessary.

Humans in the Machine

But there’s one small problem.

No matter to what extent the power grid shifts to a smarter infrastructure, there will be humans involved in the process. For that reason, the system still maintains a certain amount of risk when it comes to information security.

Cybercriminals will never tire of launching social engineering attacks that aim to take advantage of the weakest link in the chain – the gullibility of a single individual working at an energy company. Phishing emails or phone calls remain a highly effective means of fooling the target into providing access to the network’s login credentials.

If the hacker manages to gain credentials to a back-end system, it leaves the grid’s infrastructure open to attack. They could potentially reroute the supply of resources or try to turn off service for an entire region. Both results would lead to significant damage for the utility providers and their customers.

A tool of growing importance in the fight against cyber-attacks is a virtual private network or VPN. More energy companies have opted to require employees to enable this technology as a layer of protection between any internet-enabled device and the grid’s infrastructure. The concept is simple. An active VPN connection encrypts all data moving between a computer and the open internet.

To see why this is a good idea, assume a phishing scam is successful and a cybercriminal is able to hijack a local router to compromise its wi-fi network, then proceeds to intercept all incoming and outgoing messages. With the best VPN service available standing guard, the stolen messages will be unreadable thanks to the encryption and the hacker will be unable to penetrate further into the grid.

Expect the Unexpected

Given how much the power grid is dependent on natural resources and variable supplies, the U.S. government and utility companies must always be at high alert for potential disasters. In a time of crisis, customers rely on having their electricity restored as soon as possible. It’s not too much of a stretch to say the survival of civilized society might depend on it.

Judging from the aftermath of natural disasters or wars, the lack of basic services returns humans to animalistic behavior in short order.

In a best-case scenario, new software tools using artificial intelligence and machine learning will be able to alert grid operators to a potential threat or incident before the full effect is felt. This kind of smart technology is capable of mitigating the impact on grid infrastructure.

The Bottom Line

Still, humans have to be capable and ready to respond when the unexpected happens. During a time of widespread service disruptions, decisions will have to be made in terms of how to route resources. Smart systems can help to monitor the situation and provide intelligent recommendations, but that data must be used effectively by all involved.

About the Author

Will Ellis develops the guts beneath beautiful websites and can’t wait to see what the blockchain world will look like once the technology fully emerges. He invests in cryptocurrencies and studies history.


Sign up for the free insideAI News newsletter.

Speak Your Mind



  1. Thanks Will for sharing information of AI ML. Subscribed your blog for regular updates.