Comparative Analysis of Approaches for Autonomous Warehouse Stock Counting Solution

The major motive of this study is to analyse the marker based stocktaking solution [1] and compare it with the marker less AI based autonomous stocktaking solution, while understanding how these improvements in methods can be implemented. The current study focuses on a drone-based solution for warehouse environment where the stock counting process will be automated using computer vision techniques. This solution can be hosted on other robotic platforms like AMRs, AGVs, etc. The end-to-end solution with drone navigation can save a lot of time in warehouse operation of cy-cle counting by automating this routine and critical job which is labour-intensive and prone to errors due fatigue of re-peated task of manual counting. The UI of the web application is connected to the drone in the same network, where one can schedule a stock-taking operation using this platform. The focus of the solution of box counting also encompasses SKU items which are not stacked uniformly.