The mission of the engineering school at Grand Valley State University (GVSU) is to prepare graduates to meet the needs of local manufacturers. Common among the list of needs is the ability to design and implement controls systems [Adamczyk et. al., 2002]. There is also a less frequent, but highly valued ability for students able to integrate manufacturing systems. The traditional approach to satisfying this demand is to educate students in linear control systems techniques and then present a course in integrated manufacturing that includes sections on robotics, PLCs, etc. Initially we started with this approach, but quickly shifted to an approach that has been widely praised by our graduates and industry. The new sequence of three courses is: EGR 345 Dynamic Systems Modeling and Control; EGR 450 Manufacturing Control Systems, and; EGR 474 Integrated Manufacturing Systems.
EGR 345 begins with modeling continuous systems including translational, rotational, electrical and motors. This is then used to support linear controls topics such as feedback control, block diagrams, Bode and root-locus plots, and PID control. The course de-emphasizes Laplacian analysis in favor of explicit integration and numerical methods. There is an extensive laboratory component that uses a variety of basic and advanced hardware to reinforce the theoretical portion of the course. The basic equipment includes proximity sensors,
ultrasonic range sensors, accelerometers, Labview and DAQ cards, oscilloscopes, permanent magnet DC motors, ball screw slides, Allen Bradley Ultra 100 brushless servo drives, Allen Bradley Ultra 5000 Brushless servo drives (programmable in C), Vexta stepper motor drives, and Allen Bradley variable frequency drives. When the course is complete the students are capable of understanding, analyzing, designing and implementing a single degree of freedom linear control system.
EGR 450 uses Programmable Logic Controllers (PLCs) for the discrete control of manufacturing systems. Topics that are covered include sensors and actuators, Boolean system design, state system design, number systems, PLC memory structures, basic and advanced PLC functions, analog IO, PID control, and data communications. The course includes weekly laboratories that involve electrical wiring of industry standard hardware and programming. The hardware used includes Allen Bradley PLCs (PLC-5s, micrologix and SLC5s), RS-Logix programming software, miscellaneous proximity sensors, solenoid valves and push buttons. The course concludes with a major design and build project, quite often done for a local manufacturer. After this course students are ready to design and build controls systems in an industrial environment.
EGR 474 teaches the integration of a variety of hardware and software including robots, desktop CNC machines, DVT vision systems, servers (Linux), SQL databases, networking (Ethernet), Devicenet, etc. The course is taught largely as a laboratory/project course. The students are taught how to integrate the systems primarily through programming in C. During the course the students are ushered from being able to communicate with single devices through to being able to integrate the entire architecture through a database. The course concludes with a project to integrate a complex workcell to produce a simple product. After this course the students are positioned to work in advanced engineering environments planning and implementing the architectures of large scale integration projects.
The hands-on use of industrial equipment throughout all courses allows the students to understand the context of the theory and increases their interest. As a result the students are more eager to learn the theoretical fundamentals in the control sequence. This actually leads to the counterintuitive result that by doing hands-on work the students are able to learn more theory at a higher level.
EGR 345 is the first of the three course sequence and is required for all Mechanical and Manufacturing engineering students in their fifth semester. The course format is three one hour lectures, and a three hour lab each week. The course originally began in a traditional format focusing on modeling and control of linear systems using Laplacian techniques. It has since evolved to become a mature modelling course focusing on methods and topics more relevant to non-electrical engineers [Jack, 2002a]. Along with this, the laboratories were redesigned to make generous use of industrial hardware and software.
The primary shift in the theoretical part of the course was to eliminate Laplace transforms and increase the emphasis on analysis techniques using calculus and numerical methods. At the completion of the course students are quite adept at analyzing complex systems, including non-linear models. Moreover, the additional time spent solving differential equations helps to resolve fundamental mathematical deficiencies [Adamczyk et.al., 2002]. In theoretical terms most of the traditional linear controls analysis techniques are still available. In many cases the Laplace operator ‘s’ and the differential operator ‘d/dt’ are interchangeable. The topics now covered in the course are listed in Table 1. The elimination of Laplace transforms also made it possible to expand the topic list to include motion control, sensors, actuators and Analog I/O.
A list of the laboratory experiments is given below. The students use industrial hardware, software and sensors to conduct experiments in the first half of the semester (experiments 2 to 7). For example, in experiment 6 a cylinder that contained a combined spring-damper combination was investigated. The students determined the spring coefficient using forces and displacements. The damping coefficient was then determined by measuring position as a function of time using an ultrasonic range sensor. The data was collected using Labview and a data acquisition card. By the end of experiment 7, students were able to select and use sensors such as accelerometers, potentiometers, proximity sensors (photo, inductive and capacitive), ultrasonic range sensors, strobe tachometers and load cells. They were also exposed to issues of electrical wiring, noise and some simple filtering strategies.
Labs 8, 9, 11 and 12 focussed on industrial motors and controllers. The labs started with tutorials to get the students acquainted with the hardware, followed by further investigations. In the servo motor and VFD labs the students were exposed to the effects of motor control variables. They changed PID controller variables to make the systems under/overdamped or unstable. The stepper motor
At the completion of this course the students are able to analyze all of the problems that they could in the Laplace version. However, they were also able to analyze much more complex (non-linear) problems, and they had knowledge of industrial hardware.
EGR 450 is the second course in the sequence, it is required for all Manufacturing engineering students in their sixth semester. It is also a popular option for students in Mechanical engineering. The format for the course is three one hour lectures, and a three hour lab, each week. This course primarily uses Programmable Logic Controllers (PLCs) as the basis for the analysis and design of discrete event control systems [Jack, 2002b].
This type of course is uncommon and is rarely found at the university level. This is primarily because PLCs in the 1970s and 80s were very limited in complexity and application. Since then they have grown to be ubiquitous in industrial equipment. The have also become much more sophisticated and have become the subject of research in some cases [Feldmann et. al., 1999]. By offering this course, GVSU has been able to distinguish itself among local companies, and many of our graduates have become controls engineers as a result. The topics in the course are listed below.
These topics are closely tied to lab exercises. In general, labs required the students to design a program, and then debug it in the lab. The labs also help develop documenting habits and practices. The labs are listed below. Labs 1 to 6 cover fundamental logic techniques and topics that would be expected from a technician. Labs 7 and 8 focus on feedback controller implementations. Lab 9 focuses on interfacing controllers to communicate data. Labs 10 and 11 allows the students to work in small teams to develop control programs for a four station line. They must then integrate the four controllers to communicate and control the line. Finally the students integrate a vision system with a PLC in lab 12.
This course concludes with a major project that involves the design and construction of a control system to control a real process. In the summer of 2001 seven of the twelve projects were done for local companies.
The third and final course in the controls sequence is the Integrated Manufacturing Systems course, EGR 474. This course is an elective designed primarily for senior manufacturing engineering students, but it is also open to students in electrical and computer engineering. In the original format the course took the ‘grab-bag’ approach to topical coverage that is typical in other universities and texts. Since then it has evolved to focus on hands-on development. In the current format it uses two classes per week, each three hours long. Students are expected to read the text [Jack, 2002c] before each class. During the class time they then follow tutorials to familiarize themselves with the technology, and then work with the equipment at a higher level.
Students entering the course already have a knowledge of C/C++ programming, and some experience implementing computer and/or PLC based controls. Early in the course students set up a Linux computer to act as a network/web server, and a development platform. This is subsequently used to develop drivers for communication over networks, serial cables and with an SQL database [Postgres, 2002]. After this, students get exposure to low level devices such as robots, write drivers to communicate with them, and work towards developing higher level control architectures, as shown in the following topic list.
In this typical sequence, students get familiar with servers and mature operating systems in topics 1 and 2. Topics 3 to 5 deal with communications issues and device drivers. Topics 6 and 7 prepares them to deal with a device in depth and gain some expertise. Students then shared this expertise with their classmates through oral presentations and written tutorials. Finally, the project allows the students to partake in the design and implementation of a mature integrated system.
The workcell was designed and developed by the students from beginning to end. The objective the class of 2001 selected for the workcell was to mill out pen holders on 6” by 6” boards. They would be customized with a ‘S’ or ‘M’ logo, for one of two nearby football teams. The students planned the architecture for the system shown in Figure 1. They divided the work into smaller tasks, and then implemented and tested the system.
The physical layout of the system is shown in Figure 2. The cell is interconnected by a material handling system with straight track sections and turntables. The carts are controlled by ‘lock-and-gos’ which can lock to stop a cart and the propel it with a pneumatic air cylinder. On the straight track sections the lock-and-gos are in the center of the sections. The turntables have a stepper motor controller that will sense the presence of carts, rotate appropriately, and then ‘kick’ the carts to the next station. In the system the turntables use self contained controllers, but the other cylinders on the straight track sections are actuated by a Devicenet based control system.
In the design chosen by the students, an order is placed on a computer via a web browser. This then causes a wood block to be ejected from the feeder. A robot loads this into the CNC mill where the ordered pattern is cut. After this the pen holder is loaded onto a cart, and moved to the vision station where it is checked for correctness. If it is not correct the pen holder is ejected at the ‘eject’ station, and the process restarts at the feeder station. Otherwise the cart continues round to the ‘Pickup’ station where the block is given to the customer. After this the cart returns to the pickup point. This is summarized in Figure 3.
All of the software in the system was written and run under Linux, with the exception of software for programming the vision systems. The programs all communicated through the SQL database to coordinate device actions, and to track order data.
At the conclusion of this course is that students are able to develop a high end integrated manufacturing system within a single course using industrial equipment. This was not possible when using the older ‘grab-bag’ course outline, or using desktop oriented operating systems.
Most engineering curriculum focuses on a narrow region of theoretical problems and equipment that leave students unable to solve industrial problems. This paper has presented an approach, in three courses, that is theoretically mature, but leaves students able to implement common industrial control systems. This benefit is further enhanced by using industrial grade hardware and software that is commonly seen in industry. Students recognize the relevance to practice and are much more willing to learn difficult topics within this context. This also appeals to employers who are able to hire recent graduates who are quickly capable of being productive contributors. The success of the approach can be measured through the senior projects in the school of engineering. For example, in 2001 there were 5 projects done in teams of 4 to 6 students with an average industrial contribution of $20,000 for each project. The project topics were;
• A servo controlled glue deposition system - a two axis servo controlled glue over multiple paths was designed and build for a local manufacturer. The machine is now in use in production preparing cloth patterns for bus seats.
• A door beam thickness tester - a PLC based machine was designed and build for an auto parts manufacturer that would gauge door beam thickness for 100% inspection. This machine in now used in production.
• Altitude testing chamber - a chamber was purchased and temperature and pressure controls were added. This included a Labview based system, and VFD. The system is in use in the testing lab of a local avionics manufacturer.
• Combined Bending and Torsion Apparatus - this project involved the design and build of a pressure vessel that could have bending and torsion loads applied, and the stress measured. The system uses Labview for data collection, and is currently being used to support an undergraduate course.
• Die Readthrough - This was a research project for a local auto parts manufacturer who was examining the effects of die changes. The students used Labview as an integral part of their data collection and analysis system.
Feldmann, K., Colombo, A.W., Schnur, C., Stokel, T., “Specifications, Design, and Implementation of Logic Controllers Based on Colored Petri Net Models and the Standard IEC 1131 Part I: Specifications and Design”, IEEE Transactions on Controls Systems Technology, November 1999.
Hugh Jack is an Assistant Professor in the Padnos School of Engineering at Grand Valley State University. He has been teaching there since 1996 in the areas of manufacturing and controls. His research areas include, process planning, robotics and rapids prototyping. He previously taught at Ryerson Polytechnic university for 3 years. He holds a Bachelors in electrical engineering, and Masters and Doctorate in Mechanical Engineering from the University of Western Ontario.