E-Charge 2: Pioneering AI solutions for heavy transport electrification
AI-optimized charging for electric heavy transport. AI Sweden enables accurate power forecasting & smart grid use for zero-emission trucking.

In the E-Charge 2 project, AI Sweden uses advanced AI and Machine Learning (ML) to optimize charging for electric heavy transport.
Our key deliverables include comprehensive data mapping for AI training and an ML model to improve power forecasts, crucial for optimizing grid use and speeding the shift to emission-free logistics.
Collaboratively, we map data from grid companies, CPOs, and logistics operators, addressing data gaps and legal/technical hurdles, while curating datasets and establishing IT infrastructure. We lead investigations into ML's practical use for power prediction and charging optimization, employing Federated Machine Learning to ensure data privacy.
The goal is enhanced power forecasting, optimized infrastructure design for efficient grid use, and data-driven tools to accelerate investment in a resilient, zero-emission transport future.
Challenges
Electrifying heavy transport is vital for sustainability but presents key challenges. Accurately predicting power demand at charging stations—when, how often, and for how many trucks—is crucial for grid optimization and reliability.
The main hurdle is data access. Without sufficient data, forecasting grid load to balance supply/demand, integrate renewables, and optimize station use becomes difficult.
Our innovative approach involves collaboration and privacy-preserving technology. Data stays with our partners (grid companies, CPOs, logistics operators). We develop tailored forecast models for each and then securely share insights from these separate models (e.g., production, consumption, and charging need forecasts), to gain a holistic understanding. Federated machine learning can be used to improve the consumption model, predicting the power grid load, by including data from several power grid companies.
Collaboration and privacy-preserving technology enable us to optimize grid and infrastructure use, creating an efficient charging ecosystem that balances needs, costs, and availability, speeding the shift to zero-emission heavy transport.
Project purpose
AI Sweden's focus in the E-Charge 2 project is to pioneer an optimized system design for heavy electric transport charging infrastructure.
Our core mission is to:
- Deliver a comprehensive mapping and analysis of the data crucial for effective AI training
- Develop a machine learning (ML) model specifically designed to significantly improve power forecasts
By applying advanced AI and ML, we aim to show how these technologies can enhance forecasting and scenario tools for available grid capacity at charging locations. This work is fundamental to maximizing the use of charging stations, efficiently managing grid limitations, and creating an adaptive, stable charging process for the future of emission-free logistics.
Expected outcomes – Data insights & predictive power for E-mobility
Data and AI-driven understanding of demand of charging systems for heavy electric transport. A data blueprint to guide future AI development and collaboration within this field.
Development and validation of cutting-edge Machine Learning models that empower stakeholders with better tools for grid management and infrastructure planning, paving the way for transitioning into emission-free heavy transport.
Practical evaluation of privacy-preserving techniques like Federated Machine Learning, showcasing how sensitive data can be leveraged collaboratively to solve industry-wide challenges.
Facts
Funding: Vinnova (~205 MSEK whereof AI Swedens part ~2 MSEK)
Project period: The project is set to end 2027-12-31
The project's 42 partners are:
AI Sweden
Alfredssons Transport AB
Alltransport i Östergötland AB
Berndt Mattssons Åkeri AB
Börje Jönsson Åkeri AB
Chalmers Tekniska Högskola AB
Circle K Sverige AB
Dagab Inköp & Logistik AB
DFDS Logistics Services AB
Ecolink (Jula)
Erikssons Åkeri i Tomelilla AB
Falkenklev Logistik AB
GA Erikssons Åkeri
GDL AB
ICA Sverige AB
Jardlers Åkeri AB
Kilafors Industri AB
Kungliga Tekniska Högskolan
Laddbolaget i Värmland AB
Lars-Eric Bergman Transport AB
LBC - Frakt i Värmland AB
Lindholmen Science Park AB (coordination and partner)
Linköpings universitet
Lunds universitet (LTH)
LTU Business AB
Martin & Servera AB
Milence Sweden AB
Närkefrakt Ekonomisk förening
OK-Q8 AB
Rifil AB
Scania CV AB
Scania Transportlaboratorium AB
Smurfit Kappa Kraftliner Piteå AB (SMURFIT - Westrock)
Sveriges lantbruksuniversitet (SLU)
Tyréns AB
Uppsala universitet
Varberg Energi AB
Vattenfall AB
Viscando AB
Volvo Technology AB
Wibax Group AB
For more information, contact

Other energy sector projects

Simulation and visualization of Sweden's future electricity system

Future electricity grids created with federated machine learning

Federated machine learning for creating power load profiles
