[DNPP1] AI-Based Identification and Counting of Plastic Pipes in Warehouse and Outbound Operations at Dong Nai Plastic

At Dong Nai Plastic’s factories, warehouses store thousands of plastic pipes with various diameters, lengths, and types. Pipes are stacked in bundles on warehouse racks and on trucks during outbound delivery. Currently, inventory counting, inbound–outbound reconciliation, and verification of pipe quantities before trucks leave the factory are mainly performed manually or based on warehouse documents. This process is time-consuming, prone to errors, difficult to monitor in real time, and highly dependent on staff experience.

- Automatically identify and count plastic pipes from images or videos captured in the warehouse and on trucks before delivery - Identify pipe size, at minimum nominal diameter (more detailed classification if possible) - Generate quantity reports by item code to support inventory checks, inbound–outbound reconciliation, and quick verification before shipment - Reduce inventory checking time, minimize errors, and increase transparency in goods management Message to Startups: This should be viewed as a Computer Vision + Digital Workflow challenge in a real industrial environment, where solutions must be both “intelligent” (AI-driven) and “practical” (easy to use, cost-effective, and durable for factory operations). The ultimate goal is to reduce inventory counting time, minimize errors, and increase goods transparency, enabling Dong Nai Plastic’s factories to operate more efficiently and reliably at scale.

Manual counting by warehouse staff Use of warehouse inbound and outbound slips

- Manual counting by warehouse staff - Use of warehouse inbound and outbound slips

- Warehouse staff - Factory operations team

N/A

Must operate in large warehouse environments with tall racks Must handle changing lighting conditions (daytime, nighttime, natural and artificial light) Pipes on trucks may be unevenly stacked, overlapping, or misaligned Applicable devices may include: + Fixed cameras in warehouses + Smartphones or tablets + (If applicable) fixed cameras at factory gates or weighing stations Output data must be stored (file, API, or database) to support future ERP integration