Demonstration of low cost automation solutions for SME adoption of digital manufacturing

Posted on

< Project Overview >

This project addresses a common industrial concern that recent developments in digital manufacturing are unlikely to be accessible by SMEs owing to the associated capital cost of upgrading industrial computing and communication environments.

It proposes a radically different approach to the digital evolution of a manufacturing operation by focussing predominantly on non-industrial solutions to industrial automation and information challenges. It will seek to exploit very low cost commercially available technologies for mobile computing, sensing, AI and tackle the challenges associated with integrating these safely and securely into a small scale manufacturing environment.

As part of this we propose a demonstrator development which will examine issues such as automation coordination with very low cost computing, integration of low cost IoT sensing devices to supplement or enhance existing sensory systems, investigations into the real time capability of cloud based control, and the use of open source machine learning algorithms and code for process improvement. These demonstrations will provide support to an EPSRC funded project focussing on the possibilities of using very low cost automation. The project is thus divided into three mini-projects:

1) Low cost, self-managing production cell

In this mini-project we will develop a Raspberry Pi (RPi) based robotic system on which PLC logic is run. This will enable robot operations to be automated continuously, safely and reliably.

2) Self-learning 3D printing system

In this mini-project we integrate self-learning capabilities into a 3D printing (3DP) system controlled by the RPi and integrate the operations of the robotic system with the 3D Printer. Specifically, a vision system and IoT connected sensors will be introduced to extract process parameters and analyse their influence on the product quality. Data will be uploaded onto the Cloud and processed with open source machine learning algorithms to infer optimal 3DP system parameters to achieve higher quality products.

3) Assessment of real time capabilities of an IoT based low cost automation system

Building on the two projects before, this mini-project will produce potential architectural options for incorporating low cost sensing and automation pathways for SMEs, produce metrics for assessment of these options, design an experimental methodology for assessment, and demonstrate the use of the assessment methodology through interviews with SME partners.

Project Lead:

Professor Duncan McFarlane, Professor of Industrial Information Engineering at the University of Cambridge.

(Duncan also leads Digital Manufacturing on a Shoestring, a collaborative project across manufacturing SMEs and researchers, harnessing low-cost technology solutions to support growth and productivity.)

Partners:

A number of SME partners will be engaged throughout the project.

< Related Content >