Redefining Scanning with AI: Multi-Label Detection at DHL

Powerhouse AI background image
Powerhouse AI hero element

Redefining Scanning with AI: Multi-Label Detection at DHL

Powerhouse AI background image

Redefining Scanning with AI: Multi-Label Detection at DHL

Summary & Results

Facing the labor-intensive task of scanning up to seven barcodes on each of their 100,000 cartons four times a year, DHL sought a more efficient solution. Powerhouse AI transformed the tedious manual scanning process into an easy phone swipe, scanning multiple barcodes in one go. Next to this, the application guides the user through the the cycle count process and immediately highlighting discrepancies. A comparison test on 950 boxes revealed that our solution cut stocktake time by 36%, completing the task in just over 23 minutes—13 minutes faster than the traditional scanning method. In addition, the user-friendly design enabled full efficiency with less than 10 minutes of training, underscoring the solution's potential to significantly enhance warehouse operations.

Maximum productivity in <10 min of training

100% accuracy

36% faster stocktake

10x faster scanning

Logistics (healthcare)
Company Size
600,000 employees
Key Features
Multi label scanning
Key Processes
Cycle count


Our client, in accordance with a service agreement, is obligated to conduct an inventory check in their warehouse four times a year. This is no small task as their warehouse houses 100,000 cartons, all of which need to be manually scanned during each cycle count.

Each carton comes with up to seven barcodes, each needing to be scanned individually using a traditional barcode scanner. This includes the LPN, lot number, serial number, expiry date, and quantity. This is not just a tedious process, but also time-consuming.

In addition to the extensive labor required for each cycle count, the warehouse also struggles with maintaining an accuracy level. Their accuracy level is below 99%, a slight margin that nonetheless has significant implications. Any inaccuracies can lead to surplus inventory, an issue that becomes especially problematic given the space constraints within the warehouse.

In light of these challenges, our client was on the lookout for a solution that could speed-up the cycle counting process and enhance the accuracy of their inventory data. They needed a more efficient system that could streamline operations and ensure precision.


Our mission was to develop a solution that transformed the task of scanning multiple barcodes into an activity as simple as moving your phone. The result was an application designed to work in sync with your existing workflow.

The application's core function begins with the import of the cycle count schedule - either directly from the Warehouse Management System (WMS) or as an Excel upload. Once the schedule is successfully processed, the software takes on the role of a virtual guide, leading the user from one location to the next.

At each location, the user is prompted to scan a location barcode. This acts as a trigger, opening up the location for product barcode scanning. 

Once the location barcode is scanned, the user can then move the phone past the boxes on the shelf. All product barcodes within the location are automatically read and linked to that specific location in the system. The application provides immediate visual feedback for each successful scan.

When the barcode scanning at one location is complete, the user clicks "next", signaling the application to guide them to the subsequent location.

Throughout this process, the application remains vigilant for discrepancies. If detected, these inconsistencies are displayed to the user. This prompt feedback allows for immediate action, ensuring accuracy and efficiency in their warehouse operations. The solution turns the tedious task of barcode scanning into a streamlined, user-friendly experience.


Our client wanted to measure the efficacy of our scanning solution. They devised an artificial scenario within their warehouse to directly compare our technology with the traditional Zebra barcode scanner. They selected a random warehouse section containing approximately 950 boxes, to serve as the testing ground.

A company operator, experienced with the classic Zebra scanning solution, initiated the benchmark test. It took him a considerable 36 minutes and 20 seconds to scan all boxes.

The same operator, after a brief 10-minute training on our Powerhouse AI solution, took up the challenge of re-scanning the boxes. The time difference was large. The total scanning task was completed in just 23 minutes and 10 seconds, yielding a time savings of over 13 minutes, a 36% increase in efficiency!

To ascertain the ease of use, a Powerhouse AI employee familiar with the app replicated the experiment. The result? An identical timing of 23 minutes. This further validated our solution's user-friendly design as the company operator reached full efficiency in less than 10 minutes of training.

In essence, the Powerhouse AI multi-label scanning solution not only accelerates the scanning process by an impressive 36% but is also highly intuitive to use. It redefines warehouse operations, enabling companies to capitalize on efficiency and speed.

Want to eliminate manual work?