By Giorgio Sarno, Data Scientist at Stratio Automotive
The European Union's 2022 decision to ban the sales of all internal combustion engine (ICE) vehicles from 2035 has meant that the transition to alternative sources of power, such as electric, is an impending necessity for public transport operators.
It is a requirement to reduce carbon emissions and promote eco-friendly practices across the transport sector, but it requires transport operators to develop a strategy that will ensure a successful shift.
While the long-term environmental benefits of electric vehicles (EVs) may be apparent, the upfront cost of transitioning to an electric fleet requires significant capital investment.
Even with the cost of batteries and chargers starting to come down, the cost of most electric buses is still substantially higher than their ICE equivalents. The EV landscape also represents unfamiliar ground for traditional fleet managers, who will need to update their knowledge around whole life vehicle costs, battery maintenance, route management, and more.
Moreover, there are still fewer vehicle technicians qualified in electric vehicle maintenance than their ICE counterparts, making it harder for operators to recruit qualified workshop staff.
In order to successfully manage this expensive transition, fleet operators must develop a strategic understanding of the EV transition and its long-term impact on their bottom line. In order to ensure a return on their investment, transport providers will need to operate vehicles for longer periods and reduce overall fleet and labour costs. This cannot be achieved using traditional reactive and preventive approaches to maintenance.
AI-powered predictive maintenance
The adoption of predictive maintenance solutions is critical to successfully deployment of an EV fleet. By keeping vehicles out of the workshop and on the road, predictive maintenance enhances the economic viability that is needed for zero-emission vehicles’ mainstream adoption.
In fact, with a technology as new as electric buses, AI-powered predictive maintenance technology is becoming increasingly popular in the transportation industry to give operators the ability to collect and analyse vehicle data.
This not only helps to predict component failures before they cause bus breakdowns, but also provides maintenance managers with the visibility they need to ensure optimal asset management. The technology uses machine learning algorithms to analyse vehicle data, including information on energy consumption, performance, and other key metrics.
The data is then used to spot patterns that enable more efficient operational strategies and indicate possible failures, giving operators complete control over the health of their vehicles.
Pioneers in the deployment of fully electric fleets, such as Arriva Netherlands, have already leveraged this technology to monitor the state of health of their vehicles and to analyse battery performance.
Building the case for EVs
Arriva isn’t the only transport provider to have turned to predictive maintenance to increase the efficiency of its electric fleet.
Regarding the need to transition to a data-driven approach to support the deployment of zero emission vehicles, Pierre Gosset, Keolis’ executive director of the Industrial Division, said “pushing to get additional electric or hydrogen vehicles as part of a decarbonization strategy requires tele-diagnostics, real-time information, and a deep learning capability on large data sets to improve things. You need tons of information and to move from the mechanic world to the data-centric world”.
By integrating predictive solutions into the operation of EV fleets, fleet managers are able to establish the true remaining useful life (RUL) of components and use the information to intelligently schedule vehicle servicing in order to extend operational life.
This is a far more cost-efficient approach than preventive servicing where parts were replaced before they reached end of life in order to prevent breakdowns due to the higher component and labour costs of EVs.
Addressing battery life
EV batteries represent a clear example of the benefits of lengthening the lifespan of components. Accounting for around 40% of the total vehicle cost, public transport managers must attempt to extend the battery’s longevity if they hope to achieve a profitable shift to electric fleets.
Calculations need to be made to account for the degradation of the battery over time and the impact this has on range. Without this understanding and visibility, it will be impossible to plan for a smooth, efficient and cost-effective service.
Predictive battery analytics can provide an accurate, comprehensive view of the battery health evolution of an EV bus, allowing for effective route planning and charging requirements, as well as usage optimisation metrics to extend the lifespan of the vehicles.
By leveraging State of Charge (SoC) and Depth of Discharge (DoD) data, fleet managers can understand if the operation profile can be changed to maximise battery life, reducing the total cost of ownership of electric buses. This type of analysis is fundamental for an operationally successful and profitable EV fleet deployment.
Achieving lowest cost per mile
By gaining real-time, actionable insight into the internal faults of electric buses, maintenance managers can diagnose malfunctions remotely. Vehicles stay on the road for longer, maintenance becomes more predictable and less expensive and breakdowns happen less frequently, preventing costly downtime and service disruption.
Embracing predictive maintenance technology will not only smooth the transition from diesel to electric, but also sustainably spearhead fleet operators into the next generation of smart, sustainable mobility. By pushing costs down and improving reliability, public transport providers will be able to make their vehicles run for longer, serve more people, and secure a return on their considerable investment - all while meeting their climate commitments.
About the author:
Prior to his role as data scientist at Stratio where he focuses on anomaly detection and machine learning model development for predictive maintenance in the automotive industry, Giorgio founded KNOTS, a data science company concerned with deep learning, data-driven predictive analytics, and high-performance computing. He has also worked as a Data Science Lecturer at Neural Academy and Jedha Bootcamp, imparting knowledge and skills to aspiring data scientists. Giorgio holds a Master's Degree in Physics from the Università degli Studi di Torino, as well as a PhD in Philosophy in Theoretical Physics.