fbpx

Making autonomous vehicles viable

Every year the dream of sitting in a vehicle that drives from point a to point b without a driver seems ever closer. Startups, the automotive industry, universities, and more are working on the problem daily, but the reality of owning a fully automated personal vehicle is still some ways away.

Hardware and software are only a small part of the bigger picture in intelligent transportation systems. Everything from the public’s trust to artificial neural networks that analyze live traffic data has to come together.

In the past months, EyeVi has teamed up with the University of Tartu Autonomous Driving Lab who is doing experimental research on autonomous driving to validate the current capabilities of different combinations of technologies and testing out autonomous driving in real-world situations.

What are HD maps?

The capability to determine the exact position of a vehicle in real-time is one of AV operations’ biggest hurdles. Regular GPS solutions don’t suffice since the data isn’t dynamic or precise enough. For intelligent transportation systems to reach higher autonomy levels, they need high definition maps. HD maps can be a combination of datasets including objects, models, signs, and more, but at the bare minimum autonomous systems need point cloud data. Fundamentally point cloud data is a set of points in space with X, Y, and Z coordinates. On its own, the points don’t add up to much. The key is the processing.

HD maps are where EyeVi’s specialty and contribution come to the research project. EyeVi applications enable turn-key datasets for AV solutions. With EyeVi solutions, the vehicle can localize itself with high precision while integrating live data from a vast array of sensors and subsystems.

EyeVi produced HD maps were used for two routes in the research project and we are eager to continue developing revolutionary solutions in the ideal testbed for emerging technologies that is Estonia.

Autonomous Taxis

One of the case studies being researched is autonomous taxis with the partner Bolt. Bolt aims to develop Level 4 autonomy to achieve the goal of AV-s part of their platform by 2026. Bolt and other companies in the transportation sector are the keys to seeing the large-scale commercial deployment of highly automated intelligent transportation systems.

Why it all matters?

More than 20 million people suffer injuries from road traffic crashes every year, and the leading cause is distracted driving. Even if a small percentage of those injuries can be eliminated thanks to intelligent transportation systems, then it is a victory on a global scale.

 

In collaboration with Institute of Computer Science, University of Tartu