# 5G-Enabled Smart Factory Environment: Seamless Collaboration with Arm Cobots and Mobile Robot - Savonia AMK

> This project aims to automate the seamless transfer of final products from the smart factory to the packaging line.

## Overview
This article presents a detailed examination of the collaborative integration of ABB cobot, Dingo mobile robot, and Niryo NED2 arm robots within the Lucas-Nuelle mini smart factory environment. This collaboration is orchestrated by a 5G Edge server. The project aims to automate the seamless transfer of final products from the smart factory to the packaging line.

## Manufacturing and Inspection Process
Initially, a product is ordered from the smart factory. The smart factory then manufactures the requested product. At the end of the manufacturing line, a machine vision (mVision) system, integrated with the smart factory, examines the product. It checks whether the product meets the desired specifications and records this information. After inspection, the product waits to be carried either to the packaging line or the repairment line.

The mVision system plays a key role in starting the manufacturing process. It shares critical product info, like readiness and quality, with the central server using the MQTT protocol. MQTT is a lightweight and efficient messaging protocol designed for reliable communication, especially among IoT devices, with low power consumption.

## Robot Orchestration via 5G
When the Dingo mobile robot is ready in the waiting area, it communicates its status to the server through MQTT and 5G. The 5G Edge server receives the information from the mVision algorithm that the product is ready and from Dingo that it is waiting for the product; it then sends a command to the ABB cobot to pick up the ready product and put it on the Dingo.

The ABB cobot listens to the server via TCP socket protocol. It takes the ready product with the command coming from the server and places it on the Dingo mobile robot. The ABB cobot approaches the ready product with precise positioning, uses its special gripper to grab the product, brings it to the Dingo, and releases it.

## Transport and Sorting
If the product is according to the order, the Dingo mobile robot transports it to the first Niryo arm robot. Dingo’s movement toward Niryo is directed by the server through the 5G network. Using the ROS Navigation Stack ensures Dingo moves precisely while avoiding obstacles.

When Dingo approaches Niryo’s workspace, the YOLO AI model detects it and Niryo takes control of Dingo. This control parks Dingo in the right position. Here, Dingo and Niryo communicate with each other in real-time over the 5G network. After parking, Niryo uses another YOLO AI image recognition model to find the product on the Dingo, picks it up, and drops it at a certain location.

When the product is successfully taken from the Dingo, it returns to the product waiting point and informs the 5G Edge server that it is ready to carry the next product.

If the product is not correct, the mVision algorithm informs the server. In this case, the process is the same, with the only difference being that Dingo takes the product to the point for faulty products. The second Niryo robot takes control of the Dingo and picks up the faulty product.

## Conclusion
In this study, a smart factory environment equipped with robots communicating via 5G was demonstrated. By assigning initial tasks to robots in the smart factory environment, they automatically perform their tasks in harmony with each other. The success of this integrated system heavily relies on real-time communication via the Savonia 5G network, which offers fast and low-latency communications.

## Project Information
*   **Project:** 5G:llä vauhtia digitalisaatioon (REACT-EU EAKR / Pohjois-Savon liitto, A77467)

## Authors
*   Premton Canamusa, Software Developer, premton.canamusa@savonia.fi
*   Mahdi Belila, Software Developer, mahdi.belila@savonia.fi
*   Osman Torunoglu, R&D Specialist, osman.torunoglu@savonia.fi
*   Mika Leskinen, R&D Specialist, mika.leskinen@savonia.fi
*   Timo Lassila, Test Engineer, timo.lassila@savonia.fi
*   Aki Happonen, Project Manager, aki.happonen@savonia.fi