Paper ID

3198

Paper Type

short

Description

Empowered by machine learning and artificial intelligence innovations, IoT devices have become a leading driver of digital transformation. A promising approach are augmented intelligence solutions which seek to enhance human performance in complex tasks. However, there are no turn-key solutions for developing and implementing such systems. One possible avenue is to complement multi-purpose hardware with flexible AI solutions which are adapted to a given task. We illustrate the bottom-up development of a machine learning backend for an augmented intelligence system in the manufacturing sector. A wearable device equipped with highly sensitive sensors is paired with a deep convolutional neural network to monitor connector systems assembly processes in real-time. Our initial study yields promising results in an experimental environment. While this establishes the feasibility of the suggested approach, further evaluations in more complex test cases and ultimately, in a real-world assembly process have to be performed.

Share

COinS
 

Augmented Intelligence for Quality Control of Manual Assembly Processes using Industrial Wearable Systems

Empowered by machine learning and artificial intelligence innovations, IoT devices have become a leading driver of digital transformation. A promising approach are augmented intelligence solutions which seek to enhance human performance in complex tasks. However, there are no turn-key solutions for developing and implementing such systems. One possible avenue is to complement multi-purpose hardware with flexible AI solutions which are adapted to a given task. We illustrate the bottom-up development of a machine learning backend for an augmented intelligence system in the manufacturing sector. A wearable device equipped with highly sensitive sensors is paired with a deep convolutional neural network to monitor connector systems assembly processes in real-time. Our initial study yields promising results in an experimental environment. While this establishes the feasibility of the suggested approach, further evaluations in more complex test cases and ultimately, in a real-world assembly process have to be performed.