Data analysis and processing lead to eco-driving strategies

In a previous news story, we explained how JOSPEL is developing a new communication infrastructure that will make it possible to adjust the energy use in electric vehicles. One way to do this is this is through eco-driving strategies that are based on real-time analysis and in-depth data processing.

The benefit of the communication infrastructure is that it makes it possible to share information and data generated by the vehicle, and in turn develop strategies that can make electric vehicles far more flexible, autonomous and provide them with a longer range. However, to achieve this, there are several steps along the way. Earlier in the JOSPEL project, two important tasks have been to define the monitoring infrastructure that will collect the energy consumption information from the vehicle, and to describe the different eco-driving strategies that can be implemented to ensure that the vehicle is always in the best possible condition. After the completion of these tasks, the focus has turned to the overall workings of the system, i.e. describing how all the pieces will fit together and ensuring that the system makes the right decisions in time and send the proper commands to the components or subsystems involved.

Inherent limitations of electric vehicles

When discussing the workings of the system, it is important to highlight two main aspects that should not be forgotten when developing solutions for electric vehicles. Firstly, access to internet can, unfortunately, not always be taken for granted, as there are still a considerable number of spots that are not covered or have a limited capacity. Secondly, the computational power in a vehicle is far less than the one available in data centres. Naturally, these challenges have to be considered, since data processing and analysis are instrumental if you want to operate a system like the one developed in JOSPEL. Consequently, the project has assumed these challenges as pre-requisites since the very beginning, thus preparing the ICT architecture to perform well with the available resources at any given time.  With that in mind, the next logic step is to exploit the ICT architecture and provide services on top of it.

Real-time analysis and in-depth data processing

Examples of such services are eco-driving strategies that aim at maximising comfort levels in the vehicles and minimising energy waste – or in other words, taking full advantage of every Watt in the vehicle’s battery. Naturally, the results of a given policy implemented for different users and environments should not be the same. Therefore, it is very useful to collect the feedback from multiple users regarding the effectiveness of a given configuration, so as to polish them accordingly. This learning process can take place only by sharing information in the cloud through a centralised approach, in which data collection, storage and processing can be performed outside the vehicle. However, as explained before, services cannot be restricted to the availability of connectivity or to the responses after processing huge volumes of data. For that reason, JOSPEL’s implementation of eco-driving strategies is based on a double step approach, where real-time analysis is complemented by in-depth data processing.

The JOSPEL test vehicles include a component that performs real-time processing of the information and signals generated in the vehicles. This analysis is based on checking whether these signals are above or below specific threshold sets for the services, to which the vehicles subscribe. This allows for a quick reaction to any change that could make it difficult to reach the target destination – or that simply represents a waste of energy. At the same time, information about the vehicle performance is reported to the JOSPEL cloud. Here, other algorithms exploit the computational power of a data centre to dig deeper into the information and derive conclusions, identify patterns, discard those measures that do not work, and send updates to the thresholds implemented in the vehicle. The update is done in accordance with safety and security recommendations to avoid any issue that could risk the integrity of vehicle and passengers.

Once the complete infrastructure has been defined, and the eco-driving strategies that will result from it have been described, the next step is to integrate the system in the two test vehicles for extensive testing, thus ensuring that everything works the way it is supposed to.