Back
AI Applied to Logistics - Product audit in a global company
Case image

The Logistics Situation

A global leading company in the beverage industry was facing recurring logistical challenges in its distribution operations. With a high volume of daily sales, the process of verifying products on delivery trucks relied entirely on manual inspection.

In this model, an operator would check the invoices and compare them with the loaded items. This approach, besides being time-consuming, was prone to human errors, leading to rework, delivery delays, and negatively impacting operational efficiency.

The complexity and volume of the operations increased the risk of discrepancies, which undermined client’s trust.

The Emerging Challenge

In this case, the urgent need was to optimize the operation by significantly reducing the time spent on vehicle checks, and increasing the accuracy of merchandise verification.

This required the solution to be sufficiently robust to integrate itself into the existing operational workflow, while maintaining accuracy and speed in the verification process.

Another important aspect was to ensure that the new system was scalable and capable of handling demands spikes, without compromising operational security or consistency.

Mouts’ Solution

To address these challenges, Mouts implemented a computer-vision based solution, using an intelligent algorithm to automate the verification process.

The innovation eliminated the need for manual verification, replacing it with a system that automatically reads and analyzes the products.

The operation workflow was redesigned to integrate the new technology:

    • Upon arriving at a checkpoint within the distribution center, the truck positions itself in a designated area and opens the cargo compartment.
    • The system then utilizes cameras and sensors to capture images and data from the products, which are instantly processed by the algorithm.
    • The algorithm compares the captured information with the invoices, validating the cargo in seconds.

    The Results of the Solution Implemented by Mouts

    Adopting the solution brought significant improvements in various aspects of the operation:

      • Product verification, which previously demanded lengthy periods and was subject to mistakes, now takes place in a fraction of the time.
      • A 90% check time reduction allowed the company to increase its daily delivery capacity without the need to expand its fleet or workforce.
      • Loading accuracy improved significantly, eliminating errors which previously resulted in returns or customer dissatisfaction.


      This advance not only optimized internal productivity, but also strengthened the company’s image as an innovative organization committed to operational excellence. The positive impact on operational performance indicators and client satisfaction highlights the strategic value of the implemented solution.

      Technologies Utilized by Mouts

      Python

      Due to its versatility and robustness, this language was selected to be the primary developer of the algorithm.

      The combination of OpenCV and TensorFlow libraries was fundamental in achieving the desired results.

        • OpenCV: was utilized to process the captured images and videos, enabling the detection and extraction of features from the loaded products.
        • TensorFlow: with its convolutional neural networks (CNNs), it enabled the development of comprehensive models for the accurate classification and detection of objects.


        This integration between the libraries allowed the algorithm to be not only accurate but also efficient, ensuring a fast and reliable verification process.

        Azure

        The selection of Microsoft Azure as the cloud platform was crucial for ensuring the scalability and security of the solution.

          • Azure Machine Learning provided a collaborative environment for developing and training machine learning models, granting and continuous adjustments and improvements.
          • Azure Blob Storage was utilized to safely store both the captured images as well as the data generated during the verification process.


          This cloud infrastructure allowed the solution to be easily scalable, ensuring high availability even during periods of high demand.

          Count on Mouts’ expertise to find the best technological solution for your logistics scenario.