Fot. DHL / Photo: DHL press materials

Will AI-powered image recognition become a logistics standard in 5 Years? Experts say yes

Artificial intelligence-driven image analysis is poised to become a standard operation within the logistics industry in just five years, according to insights from the DHL report titled “AI-Driven Computer Vision." Part of the Trend Radar series, which keeps tabs on the most intriguing industry trends, the report predicts that AI will play a pivotal role in overseeing various aspects of logistics, from health and safety to cargo organisation.

You can read this article in 4 minutes
|

19.10.2023

Computer vision, a subset of artificial intelligence, equips computers and AI systems with the capability to extract relevant information from images, videos, and other visual data. These systems respond by taking action or offering recommendations based on the data collected or provided by algorithms.

The advancements in this field have been nothing short of remarkable. In less than a decade, AI’s ability to identify and classify objects has surged from 50% to nearly 99%, as highlighted in the “AI-Driven Computer Vision” report.

DHL cites a forecast indicating the rapid growth of the image recognition market (computer vision). In 2020, its global value stood at around USD 9.4 billion, but by 2030, it’s expected to soar to USD 41.1 billion, reflecting an average annual growth rate of 16%.

Klaus Dohrmann, Vice President and Head of Innovation and Trend Research at DHL CSI, acknowledges that logistics is already reaping the benefits of image recognition in various domains, including safety, health, and load sizing. With the development of image depth, 3D reconstruction, and technology for interpreting dark and blurry images, computer vision is poised to unlock a wealth of new opportunities and advantages for logistics companies.

The report’s authors contend that computer vision will be the linchpin for the success of logistics firms, enabling more automated, efficient, sustainable, and secure operations.

Empowering Logistics Professionals

Computer vision in logistics finds application across four core domains: health and safety, ongoing operations, asset management, and cargo management.

In the realm of health and safety, AI-driven image recognition aids in identifying potential hazards within warehouses and logistics centers, effectively minimizing risks and preventing accidents. Moreover, it can pinpoint issues such as inadequate safety attire among employees and even identify poor posture and movements, catching early signs of fatigue.

For ongoing operations, computers can utilise images to detect blockages within a facility or logistics park, analyze workflow processes, and serve as vigilant sentinels, detecting unauthorised entries into monitored areas.

Computer vision can significantly alleviate the burden on technical teams by taking on the role of monitoring and assessing the performance of logistics assets, alerting teams to any issues, errors, or anomalies that may arise.

Beyond simple monitoring of on-site activities, advanced computer vision can also automate cargo management. This technology proves invaluable in estimating load sizes and leveraging this data when planning transportation and warehouse distribution. Additionally, AI can verify cargo classification and compliance with standards and requirements, streamlining the often tedious inventory process.

AI-driven computer vision, when paired with artificial intelligence, offers extensive applications across various sectors, including retail, healthcare, and industry. It even proves invaluable in disaster response and recovery efforts.

Challenges on the Horizon

While the future of computer vision in logistics appears promising, DHL experts acknowledge several challenges and concerns surrounding its adoption.

One fundamental factor is societal acceptance, as individuals are sensitive to the idea of constant monitoring. Moreover, adhering to stringent legislation concerning personal data protection and sensitive information is imperative when implementing computer vision on a large scale. Cybersecurity also looms large on the horizon, as this technology will process critical and sensitive data.

However, ethical and legal issues are just the tip of the iceberg. Analytical machines must be meticulously programmed and trained, and the tools used for generating high-quality images must be continuously updated. Additionally, off-the-shelf solutions won’t suffice; tools must be tailored to the unique needs of each company.


This article is based on the original text authored by Michał Pakulniewicz, analyst for trans.iNFO’s Polish language service