CSEP Blogs

The need for  physical mobility remains, although we can replace physical trips with digital communication solutions. But humans have the desire for in-person interaction and “to see things in person,” which triggers both professional and personal travel. The traumatic experience of COVID taught us that we can still function as a society even when we are physically largely immobilized, but it comes at a high social cost. We need energy to enable physical transportation. The need to replace fossil fuels with renewable sources to create energy and limit or even reverse the negative consequences on the environment has become a major driver in reducing carbon emissions. This development led to regulatory frameworks for how to become a carbon-free society, which support the development of electrified powertrains and battery-based energy storage technologies, which are now mass-produced and contribute to a gradual decarbonization of the global vehicle fleet. Although the speed of this transformation is different in different parts of the world, the general trend towards sustainable mobility is, in the meantime, irreversible at a global scale. The substantial progress in AI technologies based on advanced chip design and a non-linear increase in computing power provides new capabilities both on the product side and the infrastructure side of transportation, which leads to more automation in vehicle fleet operation, design, and manufacturing processes, as well as in the delivery of supply-chain-based and customer-centric services. The human operation of mobility devices – whether on ground, in the air, or on water always comes with a social cost of human error. The social acceptance of AI-operated vehicles is linked to a safety level that needs to be orders of magnitude better compared to human-operated vehicles. This is a fundamental technical challenge that requires enormous investments in computing, networking, and sensing capabilities, both on the mobility device as well as on the supporting infrastructure. Achieving AI-based fully automated fleet operation demands substantial amounts of energy for computing to train the vehicle fleet, which needs to be provided by suitable data center capabilities based on data that is harvested through sensors, considering local conditions as well as national security regulations in the different markets. Each vehicle in the fleet needs to be equipped with sufficient local inference capabilities to master the automated transportation tasks.Preparing the transportation industry to transform towards a fully automated and fully decarbonized mobility ecosystem is a task that requires the collaboration of multiple generations of leaders and experts, bundling their collective knowledge.