Northwestern University Robert R. McCormick School of Engineering and Applied Science

Advanced Manufacturing Processes Laboratory

Operating System for Cyber-Physical Manufacturing

Motivation

The motivation for this project is to develop a cyber-physical manufacturing network that provides seamless integration, access and visibility into geographically distributed manufacturing resources. Such an operating system would leverage recent strides in control of discrete-event systems, especially in supervisor synthesis for resource allocation systems, to guarantee desired behavior of manufacturing resources while managing:

  • job flow through the system;

  • discrete optimization for scheduling manufacturing resources in such systems;

  • distributed factory automation frameworks and standards such as IEC-14699 for implementation in the form of communicating function blocks; and

  • emerging standards such as MTConnect for implementing drivers to hardware manufacturing resources.

The integration of these concepts and tools into a single system along with a strategy for permitting hierarchical synthesis will allow scalability of the cyberphysical manufacturing network. Further, the operating system will be developed for use by both, manufacturing service providers and users, thus leading to a framework that provides conceptual and implementation coupling between key entities in the network.

Description

Fig. 1: A three cell manufacturing system with shared resources on network,
which allows portable control and digital system management
Schematic

In this project, we are developing an operating system and network structure in which providers of manufacturing services and users of these services interact digitally, with a high degree of automation, to complete manufacturing transactions.

From a provider’s perspective, such a framework would safely (with respect to intellectual property, information, and hardware) expose appropriate state and capabilities/capacity information on available manufacturing resources, and accept and execute user-generated manufacturing instructions. Further, it gives the provider tools for economically managing the complexity and costs that accrue from interacting with multiple users in potentially short, one-time transactions.

From a user’s perspective, this framework provides the means to seamlessly access the desired manufacturing capacity from multiple providers in the network; manage and coordinate the logistics and flow of physical jobs in the network; and verify the services provided by each vendor in the network.

Fig. 2: Partitioning of efforts in architectural design of system
Schematic

A critical component then, is an operating system that manages both, the manufacturing ‘resources’ for the provider and the manufacturing ‘jobs/processes’ for the user, while controlling the interaction between jobs and resources.

From an overall business perspective, this framework creates the digital platform for (a) economically constructing and managing dynamic supply networks, (b) unlocking unused and inaccessible manufacturing capacity, (c) broadening the manufacturing ecosystem by allowing artisanal producers, small-manufacturers and others to participate in industrial supply chains, and (d) enabling new business models based on cyber-physical manufacturing.

In this project, a team consisting of two universities, the University of Illinois and Northwestern University, and three industrial partners, Microlution, EDM Department, and Caterpillar, aim to (1) build and deploy an operating system with network interfaces that is capable of rapidly integrating a set of manufacturing resources into a cell and accepting executables through a network interface for execution, (2) demonstrate a cyberphysical network of five operating manufacturing cells with jobs flowing between them, (3) demonstrate how new nodes can be rapidly added to the network using the operating system, (4) demonstrate that portable ‘apps’ can be developed with the operating system’s API, configurable to a node’s resources, (5) demonstrate how the resource driver specifications in the operating system can be used to support a heterogeneous hardware environment, and (6) assess the needs and identify potential apps that are needed to support the operation of cyber-physical manufacturing networks.

Approach

In this project, the research and development will be broken into the following tasks:

Task 1: Operating System for Cyberphysical Manufacturing (OSCM) kernel development: This software engineering task involves architecting and implementing of an efficient, but extendible kernel for OSCM.

Task 2: Network Operations Administration and Monitoring (NOAM) tool development: An important part of OSCM is its functioning in a cyberphysical manufacturing network (CyMaN). For CyMaN to be scalable, and to provide services that require a network-level perspective, this task identifies services that must be implemented to help users and service providers function efficiently in the network.

Task 3: Hardware Integration for Manufacturing Cells: Each of the partners in this project will take on the responsibility of bringing up at least one functional manufacturing cell operating under OSCM.

Task 4: Demonstrative Manufacturing App Development for OSCM: In this task, we will demonstrate two apps: A portable scheduling application that service providers and network users can implement and a resource finder and directory service application.

Task 5: Demonstration and Evaluation of Manufacturing in CyMaN: In this task, we will demonstrate the functioning of CMN on job-shop type tasks (single-step tasks on a single workpiece or a batch) as well as in a simple supply-chain service (multi-step processing of multiple component batches)

Sponsor

Digital Manufacturing and Design Innovation Institute (DMDII)