Unlocking the mystery: why aren't autonomous cars a reality yet?

Unlocking the mystery: why aren't autonomous cars a reality yet?

Author :
Stuart Eiffert
Stuart Eiffert, AI Engineer at Presien, explores the road ahead for automation and machine vision

Imagine a world where cars drive themselves, effortlessly navigating our streets while we sit back and relax. Despite significant advancements in robotics and AI vision, truly autonomous cars have yet to become a common sight on our streets. So why haven't we achieved it yet? 

To help answer that, we first need to understand structured vs. unstructured environments - an important concept in automation and machine vision. 

The spectrum of structured and unstructured environments

When it comes to applying AI vision technology, the distinction between structured and unstructured environments plays a crucial role. Structured environments, such as production lines, offer clear patterns and well-defined processes, making them relatively easy to integrate with AI vision systems. On the other hand, unstructured environments—think construction sites, mining operations, and bustling public spaces—pose significant challenges that require innovative solutions.

Autonomous cars occupy a middle ground between structured and unstructured environments. While there is some structure on the roads, with lanes and traffic lights guiding us, the human factor injects a dose of unpredictability. Pedestrians crossing unexpectedly, sudden lane changes, and the high stakes involved in transportation make it an incredibly complex puzzle to solve. 

So why are other AI vision solutions successfully operating right now in unstructured environments - like those used in heavy industry - while fully autonomous vehicles still face challenges?

Humans in the loop

One key difference lies in the application and the human-robot interaction (HRI) aspect. Let’s take Blindsight, developed by Presien, as an example. Blindsight is designed to augment the existing perception of workers in safety-critical environments, such as construction sites. It serves as a human-in-the-loop safety system, where the operator ultimately acts upon the system's information. In this context, some overly cautious detections can be tolerated, as it is the human operator who has the final decision-making authority.

On the other hand, autonomous cars are, by their name, fully autonomous and cannot rely on a person-in-the-loop as Blindsight does. A car driving along the highway at high speeds, without a HRI, needs to make immediate decisions based on the information it receives.  The task is much less forgiving of inaccurate detections.

Leading the way in unstructured environments

Developed specifically for the demanding world of heavy industry, Blindsight thrives in unstructured environments that leave traditional AI vision systems scratching their heads. 

Spun out of multinational construction and engineering company Laing O'Rourke, Presien’s technology was built on real sites, using data collected over more than half a decade, to ensure real-world AI performance. Fast forward to 2023 and we now have hundreds of devices operating in the field. In fact, with tens of thousands of real-world images being processed by our active learning AI system every month, our model is continuously improving and adapting.

By focusing on safety-critical applications that demand the utmost accuracy and reliability, the team has solved the most challenging problems first. So what’s next?

Expanding horizons: from safety to new frontiers

With expertise in unstructured environments, Presien is eager to apply its solutions in various domains. Logistics environments, with their added structure and predictability, provide the perfect testing ground for Presien’s capabilities. Even applications that aren't safety-critical can benefit from the advanced technology, opening up new and exciting possibilities.

Now, you might be wondering, when will those elusive autonomous cars finally hit the road? The million-dollar question! Unfortunately, without a crystal ball, no one has that exact answer just yet. While we can't predict the future, the team at Presien will continue its focus on finding new ways to put AI vision to work and to make a positive impact. 

Interested in harnessing the power of AI vision? Whether you operate in logistics, public spaces, or any other environment, Presien can provide the solutions you need to put your workplace on the path to progress. 

Put progress in plain view

Empower your team with smart insights that lift the standards of heavy industry. From safety, productivity and quality, we can build insights that change lives.