Electric vehicle charging presents multiple cybersecurity risks that smart grids alone struggle to handle.
Electric vehicles play an important role in meeting decarbonization and sustainable energy goals. The global EV market continues to grow and is expected to expand rapidly with more regulations favoring lower-emission vehicles. However, integrating EVs and their charging infrastructure into smart grids remains challenging. Their inclusion can cause numerous cybersecurity risks on both local and wider grid networks.
In a study in Scientific Reports, researchers combined a smart grid framework and artificial intelligence/machine learning algorithms to improve EV charging stations’ cybersecurity through anomaly detection and pattern recognition. Results so far have shown a cyber attack detection rate of 98.9% and the architecture can be integrated into existing EV charging infrastructure without issue.
Can algorithms provide cybersecurity for grids with EV chargers? Adapted from image used courtesy of Pexels
The Threat of Cyber Attacks on EVs Within Smart Grids
Smart grids are the backbone infrastructure for managing EVs, EV charging stations, and the wider distributed energy resources. EVs are local intermittent energy drawers from the grid, so managing them efficiently is important. However, as the number of EVs increases, cyber threats on EVs and the wider smart grid network increase because they present more potential (and often unregulated and vulnerable) entry points for hackers. Because EVs are linked to a wider information network within the grid, any potential hack can have implications across the whole smart grid network.
Remote hackers can exploit these vulnerabilities and take advantage of the weak login credentials of EV charging stations to deactivate chargers and install malware that can infect the wider network. Several common cybersecurity threats currently challenge smart grids and the wider energy sector.
Cyber-Physical Threats
Cyber-physical threats combine attacks on vulnerable digital and physical assets and can cause physical and digital damage. To inflict physical damage to the charging station, hackers can target hardware components, such as chargers or power converters, causing them to malfunction. Hackers can also overload power and communication systems, causing the charging station to shut down and potentially creating local grid instability.
Hackers can also target sensors crucial for monitoring and controlling EV systems during charging. Malfunctioning sensors can cause inaccurate readings that can affect the operational capabilities of the charging station.
Cyber-physical layers of smart grid architecture. Image used courtesy of Sharma, Rani, & Shabaz
Software Vulnerabilities
EV charging station management systems are connected to the internet, enabling energy to flow from the grid to the EVs and vice versa. These management systems contain data such as user identification, energy scheduling, and EV charging, which hackers could access by injecting structured query language and cross-site scripting.
Many EV charging stations have poor authentication systems that allow hackers to impersonate other people, leading to sensitive data theft. The software underpinning these management systems also provides a large attack surface for hackers to exploit, including making unauthorized changes to the firmware. Changes in the firmware could enable hackers to change the station’s billing systems and charging parameters.
IoEV threats
Charging stations often implement smart and coordinated charging using IoT-based systems, making smart grids and local charging networks more energy efficient. This is the internet of electric vehicles (IoEVs), where all the cars and chargers at a station are linked for more efficient charging. Charging cars at different times, depending on vehicle and grid needs, can save energy and reduce grid stress.
The IoEV system can control EV networks in real-time and contains many sensors, actuators, and communication technologies to share information. This information can be used in the wider IoT network to control the charging of EVs and manage energy consumption across charging stations.
Concept of IoEV architecture. Image used courtesy of Ahmed et al.
However, as more EVs are digitally connected to the IoEV network, it introduces more vulnerabilities to the network that could enable hackers to:
- Control energy flow between EVs and the grid, which could cause issues within the grid
- Intercept and alter the communications between the EVs and charging stations
- Infect multiple EVs with malware, either directly or by taking advantage of flaws in peer-to-peer connectivity between the vehicles
Could AI Help Prevent Cyber Attacks?
To prevent cyber attacks, artificial intelligence can be integrated into smart grids. AI and machine learning (ML) algorithms can foresee and address potential cyber attacks to increase the security of smart grid networks. These algorithms identify activities in real time by collecting data from charging stations within the network to determine which are characteristic of either normal or suspicious behavior. AI/ML can identify cyber-physical threats and software vulnerabilities and detect other anomalies within the network.
Using AI also provides a more autonomous and round-the-clock way of searching for potential threats. AI/ML algorithms will continually monitor the charging data in real-time to detect and mitigate potential cyber attacks. Alongside cybersecurity, integrating AI into smart grid networks can also help improve and optimize energy demand forecasts. Alongside AI, integrating decentralized and transparent blockchain technology could also help against cyber attacks by controlling the interactions between the grid and EVs.
