To reduce greenhouse gas emissions, the automotive industry has been accelerating its transition toward electric vehicles (EVs), while governments around the world continue to promote charging infrastructure to support this shift. With the increasing number of electric vehicles on the road, the demand for EV charging stations is expected to rise significantly. Government-backed infrastructure programs and private-sector investments are further driving the deployment of EV charging networks across cities, highways, workplaces, and residential communities.
As electric vehicles play an important role in clean transportation and low-carbon energy ecosystems, EV charging stations must become smarter to optimize energy consumption and improve charging efficiency. With IoT and AI implementation, electric vehicles, chargers, charge point operators’ cloud-based management platforms, and the grid can be connected. Key charging station data, such as charging volume, charging speed, peak usage time, energy cost, and equipment status, can be tracked and further analyzed to optimize overall efficiency and management.
To manage charging services and improve energy efficiency, AI deployment and data collection are expected to increase demand for reliable edge gateways. A powerful gateway can collect and process data, allowing operators to monitor overall usage, manage charging infrastructure, and run AI inference to support this energy transformation.
Keeping Batteries Healthy and Extending Their Life Cycle
Battery life span is one of the most important concerns for electric vehicle drivers and potential owners. Similar to mobile devices, EV batteries gradually degrade over time, and battery replacement can be costly. Therefore, maintaining battery health and extending battery life cycles are crucial tasks. In this case, AI can provide valuable support.
Deploying a gateway at the edge enables charging stations to collect driver charging pattern data, which can then be uploaded to the cloud for further AI-driven analysis and diagnosis. As AI algorithms analyze the vehicle’s charging history and battery status from the Battery Management System (BMS), smart charging systems can provide reminders and suggestions, such as choosing a suitable charging speed or stopping charging at a certain battery percentage to help protect battery life. In addition, insights into weekly or monthly driving distance can support recommended charging volumes to help avoid unnecessary overcharging.
To aggregate, process, and filter data from various sources for faster analysis in the next phase, Axiomtek recommends the ICO330 gateway. To monitor the status of energy storage systems, charging stations, and smart meters, the ICO330, powered by the Intel Atom® x6212RE or x6414RE processor, offers abundant I/O, including six isolated COM ports and three isolated 2.5GbE LAN ports. Its isolation design helps protect the gateway from electromagnetic interference and supports secure, stable data transmission.
Dynamic Pricing Based on Time and Location
By gaining insights into charging patterns, such as charging duration, peak usage hours, and station occupancy, charging service providers can use AI algorithms to set dynamic pricing based on vehicle volume, location, and time. During peak hours or in high-demand locations, pricing adjustments can help balance demand and reduce grid load. At the same time, drivers may be encouraged to charge at times that better match their schedules and help optimize charging station utilization.
To deploy AI for advanced analysis of large amounts of data received from charging stations, Axiomtek recommends the high-performance DIN-rail embedded system, the ICO520. The system features a 12th Gen Intel® Core™ i7-1265UE processor with integrated Intel® Iris® Xe Graphics, providing low-power architecture for multitasking and AI workloads. To smoothly transfer data to the cloud and control center, it provides five antenna openings and a full-size PCI Express Mini Card slot for Wi-Fi or 4G LTE modules. In addition, its M.2 slot can support faster 5G connectivity for data transmission.
Considering cybersecurity risks in EV charging infrastructure, users can also add a cybersecurity gateway to strengthen system protection. The iNA100, featuring an Intel Atom® x5-E3930/E3940 processor, can be deployed as a firewall. It provides four GbE LAN ports for connection to other gateways, and one pair of LAN bypass is built in to help prevent a single point of failure and traffic interruption. It also supports Wi-Fi, 3G, and 4G/LTE connectivity.
License Plate Identification for Charging Space Access and Payment
With the increase in electric vehicles and charging stations in urban areas, parking spaces remain limited. In some cases, charging spaces may be occupied by gasoline-powered vehicles that do not require charging. With AI deep learning and image enhancement technologies, charging stations can identify license plate numbers more accurately under various weather conditions and compare them with a whitelist. When registered electric vehicles arrive, the ground barrier can unlock automatically to allow access to the charging space. As cashless payment becomes more common, payment information can also be linked with the license plate number to make the charging process more convenient and efficient.
For this application, Axiomtek recommends the ultra-lightweight, cost-effective, fanless DIN-rail gateway, the ICO120-E3350. Based on the Intel® Celeron® processor N3350, this low-power system weighs less than 500 grams and provides essential I/O, including two GbE LAN ports and two USB 2.0 ports for cameras used in license plate identification. Users can also choose from COM, CAN, and DIO interfaces for other peripherals and ground barrier control. Designed for outdoor deployment, it supports an operating temperature range from -40°C to +70°C. An extension module providing COM, CAN, LAN, and DIO can also be added to the ICO120-E3350 to maximize connectivity and workload flexibility. The ICO120-E3350 is Microsoft Azure certified, ensuring cloud compatibility and opening access to more analytical services.
Considering the remote deployment locations of EV charging systems, Axiomtek’s out-of-band (OOB) module allows operators to monitor system status, helping reduce labor and time required for on-site maintenance. If the system freezes, it can be remotely restarted through OOB management. Since field-deployed systems may rely on solar power, power saving is another crucial requirement. The system can enter sleep mode via OOB to minimize power consumption and be turned on again when necessary. In terms of software, with services from Axiomtek’s cloud service partner Allxon, system software can be updated remotely, enabling predictive maintenance and remote management.