Watch Now

Viewpoint: Bringing sight to the supply chain’s blindspots

3D perception is taking automation to new heights

3D perception technology is able to make traditional LiDAR sensors more accurate and flexible, helping the supply chain improve automation performance. (Photo: Shutterstock)

Our supply chain is only as reliable as its weakest link. When one portion is disrupted, it sets off a domino effect that inhibits the entire system from operating fluidly. Amid unprecedented economic shocks, today’s ongoing supply chain crunch has ignited new interest in automation that can help prevent these disruptions, but the question for the industry is how to optimize efficiency without neglecting safety. 

Industry players have been working to build a more resilient system that can address the influx in demand and mitigate bottlenecks, backlogs and breakdowns. Technology that provides greater visibility into these complex, interconnected networks is critical to this effort to keep goods moving. For example, companies are bringing 3D perception solutions to market to address these acute, yet long-standing challenges. Similar to how this technology brings vision to self-driving cars, these perception solutions take the data accumulated from 3D sensors (most commonly LiDAR) to provide instant analysis which can be used across numerous components of the supply chain.

Currently, factories and warehouses use 2D cameras and single-beam LiDAR to streamline operations. But how do they stack up to the capabilities of 3D perception?

2D images and videos can be valuable solutions for a wide variety of challenges, especially when ocularcentric data is required (such as label damage detection or barcode scanning). However, 2D cameras can only detect and monitor objects in calm weather conditions, are prone to blindspots, and struggle to track across multiple cameras at once. This is fine indoors for a singular use-case, but these limitations prevent it from broad applications, like tracking, monitoring, and analyzing a large number of objects outdoors. 

Single-beam LiDAR solutions are used in factories to track the movement of people, vehicles, and robotics, as well as anticipate potential collisions. While break-beam sensors can provide more reliable safety details than 2D cameras, these systems are flawed because they are not built to distinguish and track objects. This approach can monitor and predict collisions on one plane but cannot track objects in all directions. To fully capture an environment and alleviate safety concerns, LiDAR sensors must be installed throughout a warehouse, including on the sides and ceilings, to get a stronger vantage point. 

Increasing efficiency while lowering cost

Over the past few years, the cost and computational requirements of LiDAR and other 3D sensors have lowered significantly. This decrease has led to growth in the adoption of autonomous mobility beyond the AV industry and into industries like logistics.

This transition has already unlocked advancements in automating truck loading and finished vehicle logistics, allowing for more consistent and efficient operations. This is mainly because placing the hardware on infrastructure makes it possible for thousands of objects (both static and in motion) to be managed with only a few sensors. With recent advancements in sensor application, automation is more versatile than ever before. 

Remote automation alleviates concerns regarding safety and accuracy all while significantly reducing installation and operating costs. For example, BMW is able to automate hundreds of vehicles at once from multiple vantage points; the solution is able to see around corners and predict trajectories while eliminating blind spots within a margin of error of 4 cm. These insights provide a broad understanding of the environment to reduce collisions, create a more reliable process, and increase operational efficiency within automotive logistics.

By placing sensors on infrastructure rather than on the vehicles themselves, OEMs will streamline their processes and become more cost efficient because each individual vehicle does not need to be equipped with sensors. Furthermore, the system operates fully autonomously, alleviating concerns of human error.  

While this is a powerful use case, it’s only one example of what’s possible with 3D perception. This same technology has the power to transform a wide range of applications and operations within the supply chain including the automation of trucking yards, ports, logistics hubs and more.

Creating safer spaces 

Emerging technologies cannot be considered successful if they don’t make spaces and touchpoints within the supply chain more secure. The complex system of highways, railroads, waterways and flight paths that are integral to the transport of goods must function effectively, efficiently and – most importantly – safely. Applications within the plane, train, and automobile industries have already begun demonstrating how 3D perception can be transformative, protecting cargo and assets while improving safety for workers and reducing the chances of collision and injury.

According to the U.S. Department of Transportation’s Federal Railroad Administration, 23,000 incidents have occurred from train collisions. This resulted in 2,700 fatalities and 9,500 injuries between 2018 and 2022. Railroad systems and technology integrator, Herzog, has implemented an automated obstacle detection and warning system aimed at making the interface between the railroad and public safer. Herzog’s solution utilizes a multi-sensor hardware platform and 3D perception engine to track and classify objects. Coupled with business intelligence and edge detection, the system is able to analyze the data it receives in real-time and alert the appropriate entities in advance to reduce the speed prior to collision. The solution can detect humans, vehicles, bicycles, and other foreign objects, enabling it to provide unparalleled safety insights.

Beyond railways, damage caused during aircraft ground movements have resulted in about 250,000 injuries annually, with reparations that can cost airlines upwards of $10 billion each year. Evitado, a company working to automate airside operations, has deployed 3D-based solutions at airports and within hangars to aid in the movement of aircraft for increased safety and efficiency. By installing 3D sensors on the aircraft, equipment, and infrastructure, the system can detect objects, people, and obstacles in real-time, alerting control managers if an airplane is approaching an object that could cause serious damage prior to collision. As a result of the solution’s real-time accuracy, it is also highly adaptable and can make those quick decisions regardless of environment. 

Better logistics for all 

Tapping into the power of 3D perception equips logistics companies with one of the most valuable assets in today’s economy: information. With a greater insight into facility operations and the surrounding environment, companies can ensure that they are maximizing their space, labor pool, and time. 

And warehouse automation and object detection is just the start. As the use of this technology becomes more pervasive, 3D perception will also support new use cases like strategically automating the loading and unloading of containers in truckyards and ports to create a more efficient and environmentally sound logistics system.  

Greater technological innovation will improve the accuracy of supply chain systems, automate repetitive processes, lower incident liability, and save tremendous amounts of money. This will alleviate headaches for companies and keep the supply chain stable, efficient, and safe. 

About the author

HanBin Lee is CEO & co-founder of Seoul Robotics. Based in Seoul, Korea, and with offices in Germany and the U.S., Seoul Robotics develops 3D perception software with a specialty in LiDAR sensors for automotive, factory and city infrastructure.

HanBin Lee, contributor

HanBin Lee is CEO & co-founder of Seoul Robotics. Based in Seoul, Korea, and with offices in Germany and the U.S., Seoul Robotics develops 3D perception software with a specialty in LiDAR sensors for automotive, factory and city infrastructure.