Whether you are planning to purchase new lab equipment or to start a new control / mechatronics lab; there are 2 important topics that you will need to focus on for every lab. purchase:
- What is the main goal of the system(s) that will be used in your lab
- What equipments (hard/software, services) will be needed for the system(s) to work properly
This will be a 3 episode blog and in every episode we would like to assist in answering the above 2 questions for the specific scenario of setting up either a control laboratory, mechatronics laboratory or robotics laboratory solutions.
In this episode we will be focusing on the control lab scenario first. In the other episodes we will focus on mechatronics and the robotics lab scenarios as well. One remark, although a control lab and a mechatronic lab sounds similar, building a lab solution for pure controls education differs a lot compared to building a new mechatronics or robotics education laboratory.
Let’s start our discussion with the system requirements first and then the necessary equipment requirements for building a controls lab solution.
A-Z Info on Setting up a Controls Lab
A control lab -or more accurately a feedback control systems lab- is a fundamental hands-on experiment lab required for most engineering education departments. This lab is included within the list of lab requirements in accreditation of engineering programs for topics of “engineering practice” and “making judgements” enlisted (*1)by agencies such as ENAEE, ABET or similar. Often these labs are formed under electrical, mechanical, robotics and control engineering departments but the concept of automatic control is also common among other engineering disciplines such as biomedical, chemical, computer, and industrial engineering and so on.
The primary purpose and focus of a control systems lab is to provide different hands-on industry related examples to teach feedback controlled systems in a practical, learn-by-doing fashion. According to the type of engineering discipline, the application example varies from low complexity fluid-level control experiments (mostly seen in mechanical or industrial engineering departments) to more complex multi-axis position & torque control examples that can be seen in robotics and control engineering syllabus.
Whatever the type of the experiment is, the main GOAL is always similar:
Main GOAL of a Control Lab (Experiment): Provide a repeatable hands-on experience for students to observe and apply underlying principal(s) of feedback controlled systems.
By combining the design and problem-solving skills of controls theory along with mechanical structures, sensors, actuators and most importantly control algorithm coding, a remarkable skill set will be acquired by students through the practical application of these laboratory experiments.
In general practice, we mostly see 3 different methods while executing the control lab work. These are:
- Software Only Method, ie. Simulation based Lab experiments: There are many popular engineering software, where popular experiments such as Coupled Water Tanks or Linear Pendulum(*2) are modelled computationally and a co-simulation could be run on these models. Altair Activate®(*3) is an example of engineering model-based design software. It uses either Modelica® based system simulations or an FAE based digital-twin model such as Ball Balancing Table experiment(*4) built with MotionSolve®.
- Do-it-yourself based Hardware & Software Building Method (DIY Method): As the name implies, in this method students are led to a well known experiment system and asked to physically build a replica of this well known system. There are some good online tutorials (such as this online levitation experiment(*5)) mostly using low-cost devices such as Arduino™ and sometimes 3D-printed parts as well. Once and if successfully they complete the building procedure, then they can be used to run different levels of experiments as well. It is good for interested students especially with the Maker interest. However, not all students are interested in building things from scratch, hence this method will not help 100% of the students taking the class.
- Commercial Off The Shelf Hardware & Software Method (COTS System Method): This is the well-known and safest method of building a lab. It can be considered as outsourcing the building and debugging process to an external party and executing the lab work directly on the completed, tested and verified products. This method nowadays is also divided in 2 location options:
- The classical buy, install locally and operate/maintain locally option
- The new option of remote lab execution through online Remote Lab services such as ACROME’s Remote Lab service (*6)
Each method has its pros and cons. In this blog we will not compare the methods. Rather we will pick the COTS System Method and local lab installation option (Method 3-a). Readers should keep in mind that other methods -except the Remote Lab- such as the DIY Method should also have similar requirements to install, operate and maintain the labs.
Next, let’s look at what equipment (hardware/software, services) are needed for building a Controls Lab properly.
Equipment Needed for building a Control Lab (checklist): Below is a shortened list of the crucial components which are needed when building a new controls lab:
- Physical Space (required): Although it sounds very basic, one of the most problematic areas is the physical space requirement. It is the space where lab systems will be installed and students will occupy during the lab hours. Depending on the number of systems installed, their physical sizes and how many students are present in the lab, some physical space should be allocated.
According to our experience, and also based on our desktop sized hardware, at least 3 m.2 (32 ft2) should be allocated per experiment station.
Note: Due to Covid-19 pandemic, the social distance requirements should also be considered. Above number is a suggestion for minimum placement, it doesn’t contain legal social-distance requirements of countries.
- Electrical Power, Lighting and Cabling (required): Another basic but very important requirement. Every control experiment requires electrical power to operate. Based on the type of the experiment system, the power requirement varies from 50 W up to 10 kWs. Again based on the number of experiments and the mixture of the experiment system types power budget should be calculated and an electrical infrastructure should be laid accordingly.
Additionally, there should be sufficient room illumination for the people to work in the lab.
Another important but mostly omitted topic is the cabling. In general, each experiment system may require at least 2x grid power (1 for the experiment system and 1 or 2 for the laptop/PC). So, providing outlet power to each experiment table and splitting it with a power splitter plug/cord is an important and crucial topic.
Finally, overall power consumption of the lab should also be considered as this will affect the electrical bill of the department/faculty.
- Air Conditioning (optional but most probably needed)
- (Video) Surveillance and Access Control System (optional):
- Experiment System(s) (required): This is the topic where most time is spent.
Depending on the discipline, the experiment system(s) should be selected.
For example, if the control lab is built for industrial engineering departments, then the experiment systems should be selected with more industrial components such as PLCs, industrial motors, sensors etc.
If the lab is built for mechanical engineering departments, then the components might be more low-level mechanical items such as pneumatic systems, mechanical converters, COTS MCU boards etc.
If the lab is built for electrical engineering departments, then the components might be more electronics oriented such as electrical motors, power converters, analogue controllers etc.
Also depending on the number of students to work in the lab, and student grouping, the required number of experiment systems should be determined.
Despite the difference in the components used, still most of the different control experiment systems might be used interchangeably among different disciplines. Most importantly, the systems should function properly and support the educational goals, and avoid time spent on non educational preparation and debugging tasks.
Once the type of the experiment systems are decided, then the next topic to concentrate would be selection of the vendors. There are multiple topics to focus, but we have seen the following 3 topics as the most important ones:
- Components used in the experiment system. Depending on the educational goal, the components could be selected either easy-to-use & easy-to-find ones or high-end but difficult to find/replace ones. It is also a matter of price vs. performance.
- How open the system for modifications / altering.
- What are the list of experiments that can be taught and if the companion documentation / software would support the educational goals.
Above questions are heavy and they need some time & research to answer. Best way to quickly iterate these questions would be arranging a call with the sales representatives and directly asking the questions.
- Computers (PC/Laptop) (required): Control system experiments explicitly require a computer for displaying the measurements, change the parameters, and in some cases run the control algorithm as well -as with PC based USB/PCI etc. DAQ card systems (also known as PC-based controller(*7)). In best case, the control algorithm runs on a dedicated real-time controller and the PC connects to this controller as an external device to display the system’s performance. However, PC may also be responsible for changing the parameters of the controller and in most desired case it is used for reprogramming the controller with new / advanced control algorithms. This is valid with open-source systems such as ACROME’s experiment systems with Pyhton or similar engineering software support. One major question in this topic is the performance specifications or the model / brand of the PC to be used. One thing to note that the performance requirement of the PC differs heavily on where the control algorithm is running. In case the control algorithm runs on an external real-time controller, then the performance requirement of the PC should be low-to-moderate depending on the richness of the GUI objects only. However, if the control algorithm is running on the PC (as in PC-Based Controls case), then the PC performance heavily affects the control performance, therefore the PC should be as performant as possible or at least achieve some level of performance.
In ACROME, we assist on PC selection of our customers. We can provide a list of suitable PCs upon our customer’s request. We also provide remote assistance on installing the engineering software to our customer’s own PC(s).
- Engineering Software (required):
As part of the computer item, the PC should run a specialized software to control or communicate with/reprogram the real-time controller. The engineering software handles major communication tasks and in most cases (if not all) it provides a Graphical User Interfaces for executing the experiment. There are 2 different engineering software options for any control experiment system:
- Vendor Provided (ready-to-use & non-modifiable) Application.
In this option, users/students interact with the experiment system through this application to observe measurements, interact with provided parameters, use save - load functions to write data etc. The functionality of such applications are defined and limited by the equipment vendor. It is mostly not possible to modify this application. - Open-Source Codes for well-known programming environments:
In this option, again equipment vendors provide some source codes to run on popular programming languages such as Matlab ®, NI LabVIEW® , Activate ® etc. The source codes may come very simple or may also contain a Getting Started like rich user interface objects such as below image.
Good thing about this option is that users/students can modify the existing codes for their own usage. It is also possible to try very different control algorithms which may give many other possibilities especially for the graduate and research applications.
- Training the trainer or the students on how to use the experiment systems (required)
- Courseware (required): A decent control experiment system should come with a predefined lab. manual, where teaching assistants/students can follow itss chapters to execute their lab work. It is always better to have an open-source documentation, so that the instructor/lab assistants can modify, split or add more contents to the existing documentation and share it among the students freely.
- Lab. Assistant (required)
- Maintenance Activities (required): Even though most good control experiment systems are self-contained and having some protection against mis-usage, still there is always a chance to damage the system. Also, some experiment systems (those having electrical motors or similar moving parts) may need periodical maintenance for keeping them in good shape. Overall, this is an important task that should be executed periodically and needs some work-force to do so.
- Presentation Equipment (optional)
- Internet Connection (optional)
- Lab. Management Software (optional):
In case there isn’t a lab. assistant involved, then it can be daunting to grade the students and also maintain the systems. The Lab. Management Software (LMS) can automate some tasks such as software based assessments and also running self-tests on the Lab. systems. It will be pretty handy to have a full-fledged LMS in the lab, but it is not a strict requirement.
These are the main topics that we have faced questions about building a control lab. As ACROME, we try to minimize your efforts on building your lab and maximize the true outcome, focusing on “the” most valuable teaching concepts and bringing the theoretical information into real life with the best experience.
Feel free to contact us if you think any additional topic should be added to this list.
References:
1* EUR-ACE® Framework Standards and Guidelines
2* Simulation and Animation of a Linear and Nonlinear Pendulum Model
3* Altair® Activate® Multi-disciplinary System Simulation Software
4* Ball Balancing Table Mechatronics Education Kit
5* Magnetically Levitated Ball with MATLAB and Arduino
6* ACROME Remote Lab - Accessible online hands-on experiments for control and mechatronics
7* PC-Based Controls vs. PLC-Based Controls for Machine Automation
Author
