John Anthony Rossiter- Department of Automatic Control and Systems Engineering, University of Sheffield, Sheffield, United Kingdom
This review summarises recent thinking in the academic control community on the future of control as a topic and thus on the design and focus of control courses at university. It is notable that the current thinking is quite controversial and significantly at odds with traditional practice, and thus implementing such changes will require substantial effort and will from the community.
1 Introduction
In recent years the international control community, supported by the IFAC and IEEE Technical Committees on Control1 and contol systems society (CSS), felt it was timely to have some active reflection on the control curriculum and delivery in Universities (Antsaklis et al., 1999; Murray, 2003; Dormido, 2004; Rossiter et al., 2018; Rossiter et al., 2020; Rossiter et al., 2023). What is particularly significant is the growing awareness that despite society at large and our use of technology changing vastly over the past 30 years (in essence following the advent of micro-computing, mobile phones, laptops, etc.), the control curriculum we teach has changed very little. Moreover, it is becoming increasingly evident that the module design and delivery which was necessary for a pre-computing age, was really rather dull and uninspiring to the modern generation, and thus not preparing them well for the careers they were going to enter. In summary, the need to modernise is increasingly urgent.
This mini-review will give a rapid overview of the surveys and discussions carried out by the community in recent years finishing with a summary of the current thinking and priorities. In essence this reduces to a call to arms for each of us to begin instigating change within our own institutions. Section 2 focuses on provision and design of Learning and Teaching (L& T) resources and Section 3 focuses more on curriculum content1.
2 Learning and Teaching Resources and Course Delivery
2.1 Historical Background
Traditional control courses and associated textbooks (Dorf and Bishop, 2021) have centred around mathematical algorithms and techniques for analysing and designing system behaviours, for both open and closed-loop. Many of the traditional techniques discussed and commonly assessed in control courses, assumed there was no ready access to computing, so deployed insight, clever graphs and other tricks to infer expected behaviour from some simple analysis which was amenable to pen and paper computations. Historically, laboratory access was limited and expensive and thus, while valuable, often only comprised a relatively small time component of a course.
2.2 Potential Released by Technology
With the rise of computing capacity several core changes have occured so that, even assuming a conventional curriculum, the design and delivery of all engineering courses can be significantly augmented; this paper focuses on control where some important examples are summarised here (Rossiter et al., 2018).
2.2.1 Micro-computing
The advent of cheap micro-computing means that laboratory provision can be much cheaper and more accessible, thus providing opportunities for an increased presence in the curriculum. A notable innovation that is beginning to grow in popularity in the past few years is the concept of take home laboratories (Oliveira and Hedengren, 2019; Rossiter et al., 2019; Oliveira et al., 2020; Yerolla and Besta, 2021), that is cheap (often as little as $30 per kit) and transportable mini-kits that students can borrow for many weeks and use for independent learning and investigation. Having 24/7 access to real hardware with rapid runtimes and driven by a laptop allows students to get a real feel for authentic issues. Indeed this can also inspire them and thus gain more enthusiasm for a topic.
Even in the case of more expensive and larger kits that would only be available in University laboratories (Quanser, 2022), these have become notably cheaper per unit and thus available in larger quantities.
2.2.2 The World Wide Web and Remote Access Laboratories
The creation and development of the world wide web provides opportunities for 24/7 access to both information and in some cases, real hardware. There is a recognition that University timetables limit the times that students can be physically present with expensive laboratory hardware, but allowing access via the web, potentially opens up the timetable to the full day and weekends.
A large number of remote access activities have been publicised in the control community, for example (Brinson, 2015; de la Torre et al., 2019; de la Torre et al., 2020; Egerstedt, 2022) and clearly this is just a small subset of those available. A core point to make here is that such laboratories are likely to be far more authentic and flexible than take home kits, thus providing students with the potential to further increase their insight and understanding of important engineering challenges and solutions.
Of course, there also some challenges for teaching staff in that: 1) as the numbers of kit are limited, efficient queueing of student access is needed, so access is still not totally free and 2) substantial local expertise is needed to set up the software to enable efficient, reliable and effective web access.
2.2.3 Virtual Laboratories and Interactive Tools
Along the same lines as the discussions above, the increasing power of modern computers means that student access to realistic model simulations and other interactive tools are now available through personal laptops, either as a local file running on cheap software (or available through a university license) or through a web server (Guzmán et al., 2013; Guzmán et al., 2016; Heradio et al., 2016; Rossiter, 2017; de la Torre et al., 2020).
A number of points can be made as an encouragement to teaching staff to consider deploying such virtual laboratories and interactive resources in their teaching.
1. Being software based, these may allow an unlimited number of students (in the author’s case a typical class size is around 400) to access the activity simultaneously. This also allows their use for interactive segments during lecture slots.
2. With suitably visual affects, these allow students to engage with core concepts and learning very cheaply, and thus to optimise the use of their time on actual hardware. Also simulation times can be instantaneous or much faster than real-time!
3. Modern coding tools mean that effective GUIs/activities can be coded in a half-day or less, thus cheap on staff preparation time.
2.2.4 Using Computers for Assessment
The recent COVID pandemic forced many academics’ hands in adapting the way they assess, most notably the move away from closed-book exams to open-book assessment with students sitting the test at home on their computers. An interesting discussion to be had over the next few years is whether staff feel this new model has some advantages and should be retained; anecdotal evidence in the author’s institution is that many staff liked the new model and may wish to retain it, notwithstanding issues with handling potential unfair means.
It is worth noting that there has been a quiet trend (Lynch and Becerra, 2011; Rossiter, 2011; Rossiter, 2022b) within the community for many years proposing that assessment of control computations using paper and pen exercises is somewhat ludicrous; no-one would do it this way in a job and number crunching should not be a university level assessment. We should give the students access to a computer to do the number crunching and check they can make the appropriate decisions on what are appropriate computations and design steps and indeed spot obvious errors and so forth. The author uses threshold assessment (Rossiter, 2022b), thus worth pass/fail marks only, to ensure students have base level competence in core analysis tools and this is very fast to mark using computer quiz engines.
2.2.5 Online Courses and Resources
Another area which has undergone rapid evolution in recent years, but one could argue, certainly for control topics, is developing in more of an ad hoc sense rather than systematically, is the concept of online courses and learning resources (Albertos, 2017; Rossiter, 2022a; Douglas, 2022; Egerstedt, 2022; Khan, 2022).
There is no doubt that there is a plethora of superb online resources which students can use to learn from, but a simple search on the web could return thousands of options and leave the user confused. Consequently, a current urgent project within the community (Serbezov et al., 2022) is to collate the available resources and disseminate these in a digestible manner (Douglas, 2022).
Perhaps, what is more pressing for the academic community, is the need to share resources Serbezov et al. (2022) in a manner that allows each of us, and our students, to benefit from the excellent resources others have developed. Thus the encouragement to share with suitable creative commons licenses is equally important.
2.3 COVID, Student Expectations and Summary
The recent COVID pandemic is likely to have speeded up the transition to new ways of learning and delivering learning in many institutions, especially the increasing use of online resources, lectures and assessment. The brief examples of this section serve to illustrate that much good practice already exists which staff can therefore adopt and apply relatively rapidly.
What perhaps has not been discussed, and indeed there is little space for the issue here, is the concept of student expectations and the modern student. Certainly anecdotal evidence is fairly strong that the students of today appreciate and expect different support and resources to those of students even just 10 years older. If we are to engage these modern students, we need to meet them in the right place, and it is apparent that they have grown up with digital technology and thus expect L&T methods at university to make extensive use of such technology. Indeed, as a minimum, they may expect all the resources to be available via their mobile phones.
3 Curriculum Content, Design and Delivery
The previous section has focussed largely on L&T resources and accessibility, but sitting alongside this is the core content and delivery of a first course in control. It has become apparent from recent work (Murray et al., 2004; Rossiter et al., 2021; Rossiter et al., 2023) that two fundamental changes are timely:
1. A first course needs to cover a wide range of scenarios, certainly far beyond traditional engineering.
2. A first course should be less concerned with mathematical elegance and proof and focus more on application of the core concepts.
Some interesting exemplar case studies from three varied international institutions are available in (Rossiter et al., 2023) and we hope more will follow from the respective IEEE/IFAC Technical committees in the very near future.
3.1 Applications of Control
A historical control course would likely have focussed predominantly on mechanical, electrical and chemical engineering examples such as suspension systems, tanks, heat exchangers and motors. However, there are changes in society which are pertinent. First engineers are less likely to focus on a single discipline and employers expect them to be multi-disciplinary and moreover engage in life long learning to diversify and extend their skills as required. Secondly, there is an increasing awareness (Murray, 2003) of the prevalence of feedback loops and the need for control in a wide variety of application areas such as crop growth and irrigation (Cabrera et al., 2021), solar energy (Satue et al., 2021), modelling and control of disease (Estigarribia et al., 2021), autonomous bikes (Persson et al., 2021), underwater vehicles (Rentzow et al., 2021) and indeed this list could easily be expanded and broadened far more.
A first course in control needs to be multi-disciplinary and expose students to the huge variety of potential applications and indeed, potential benefits to society, of applying feedback effectively. If we accept that university is only the first step in a life long journey of learning, then it is less important to teach students lots and lots of dry information which they can easily pick up as required. Rather we need to focus on enthusing them and exposing them to core concepts and principles, so they are motivated to engage and learn. Moreover, most institutions will have higher level courses in years three and four where students can specialise and thus engage with greater technical depth and detail.
3.2 Reducing the Focus on Mathematics
One of the controversial points in the recent international survey (Rossiter et al., 2020) was the recognition that our historical emphasis on treating a first control course like an applied mathematics course is probably not appropriate in the 21st century. In general terms, apart from a small minority, historical graduates remember just that, control = mathematics, and they have little empathy or understanding of what the topic is really about? Thus, as teachers, we have failed them.
This is not to say that mathematical precision and rigour is not important, but we have to decide its place and priority. Most graduate engineers will not be control experts, they will not need to do detailed loop analysis, understand root-loci or indeed state-space methods. However, they will need to understand what a feedback loop is, why it is important and have some understanding of the links between behaviours and tuning? Hence, we need to propose a first course that focuses on the core principles and concepts such as: modelling and behaviours, uncertainty, performance measures, the role of feedback, simple PI designs and interesting case studies/laboratories; it is noted that current discussions (Rossiter et al., 2023) may end up proposing even more drastic changes, focussing on a range of modern applications and potential usage rather than on traditional behaviour analysis.
To summarise however, while some mathematics is important, we should de-emphasise that to ensure students finish the course believing they have learnt control and not mathematics.
3.3 Exploiting Software Tools and Virtual Labs
As a throw away and discussed in Section 2.2.4, given the plethora of computing tools now available, the author believes that it helps reduce the emphasis on tedious mathematical computations and number crunching if examiners exploit computing tools (Rossiter et al., 2008) for the mathematical computations. Thus students can focus their time and effort on understanding principles and applying these to interesting applications Egerstedt, (2022); Taylor et al. (2013); Park et al. (2020).
Moreover, as discussed in Section 2, the advances in technology allow staff to adopt lots of interesting and interactive laboratory activities which bring the topic to life in a way which attendance at one or two brief hardware laboratory sessions could not. These activities, be they virtual labs, take home labs, remote labs or indeed other activities are readily available and typically low cost. Moreover, they can be used in conjunction with concepts such as threshold assessment (Rossiter, 2022b) to reduce the implied marking and assessment burden on both staff and students so the focus is on enjoying learning.
Modern students expect all their resources and much of their assessment to be accessible online. It is straightforward to do this for nearly all aspects of the learning delivery and thus is a win-win if this means students also engage better.
4 Conclusion
Universities can be rather slow to change, partially driven by legal requirements which limit the time scales for radical changes in the curriculum, but also natural inertia in staff; we have always done it this way! Of course, researchers know we cannot stand still and education likewise needs to modernise and move forward to ensure we provide the engineering graduates needed by industry (e.g. https://accreditation.org/explore-accreditation/accords/washington-accord).
Recent community wide surveys (Rossiter et al., 2018; Rossiter et al., 2020; Rossiter et al., 2023) have made it clear that both the content and delivery of control courses needs modernising in many institutions. We need a concerted effort from all academics to push forward these changes, ensuring that:
• Students are enthused by a first course in control and seek to study more advanced options.
• Students are adequately prepared for the hugely diverse problems that face them in modern industry.
• Students develop an awareness that control concepts are far more broadly applicable and useful than in just conventional heavy engineering.
• Our teaching methods and resources are tailored to the students and context of today, not those of yesteryear.
Author Contributions
The author JR has written this review paper alone, but as noted in the acknowledgements it draws substantially on collaborative work with numerous international colleagues.
Conflict of Interest
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher’s Note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Acknowledgments
The work and views summarised in this paper are a combined effort of many international colleagues, most notably the IFAC and IEEE Technical Committees on control education and working groups of the IEEE CSS.
Footnotes
1http://control-education.ieeecss.org/ and https://tc.ifac-control.org/9/4
References
Albertos, P. (2017). Mooc in Dynamics and Control. Available at: http://personales.upv.es/palberto/
Antsaklis, P., Basar, T., DeCarlo, R., McClamroch, N., Spong, M., and Yurkovich, S. (1999). Report on the Nsf/css Workshop on New Directions in Control Engineering Education. IEEE Control Syst. 19, 53–58. doi:10.1109/MCS.1999.793442
Brinson, J. R. (2015). Learning Outcome Achievement in Non-Traditional (Virtual and Remote) versus Traditional (Hands-on) Laboratories: A Review of the Empirical Research. Comput. Educ. 87, 218–237. doi:10.1016/j.compedu.2015.07.003
Cabrera, J. A., Pedrasa, J. R., Radanielson, A. M., and Aswani, A. (2021). “Mechanistic Crop Growth Model Predictive Control for Precision Irrigation in Rice,” in European Control Conference. doi:10.23919/ecc54610.2021.9654920
de la Torre, L., Chacon, J., Chaos, D., Dormido, S., and Sánchez, J. (2019). A Master Course on Automatic Control with Remote Labs. IFAC-PapersOnLine 52, 48–49. doi:10.1016/j.ifacol.2019.08.122
de la Torre, L., Neustock, L. T., Herring, G. K., Chacon, J., Clemente, F. J. G., and Hesselink, L. (2020). Automatic Generation and Easy Deployment of Digitized Laboratories. IEEE Trans. Ind. Inf. 16, 7328–7337. doi:10.1109/TII.2020.2977113
Dorf, R., and Bishop, R. (2021). Modern Control Systems. 14th edition. Pearson. ISBN 978-1292422374.
Dormido, S. (2004). Control Learning: Present and Future. Annu. Rev. Control 28, 115–135. doi:10.1016/j.arcontrol.2003.12.002
Douglas, B. (2022). Resourcium. Available at: https://resourcium.org/
Egerstedt, M. (2022). Mooc on Control of Mobile Robots. Available at: https://www.coursera.org/course/conrob
Estigarribia, P. E. P., Bliman, P., and Schaerer, C. E. (2021). Modelling and Control of Mendelian and Maternal Inheritance for Biological Control of Dengue Vectors. European Control Conference.
Guzmán, J., Costa-Castelló, R., Dormido, S., and Berenguel, M. (2016). An Interactivity-Based Methodology to Support Control Education: How to Teach and Learn Using Simple Interactive Tools. IEEE Control Syst. Mag. 36, 63–76. doi:10.1109/MCS.2015.2495092
Guzmán, J. L., Dormido, S., and Berenguel, M. (2013). Interactivity in Education: an Experience in the Automatic Control Field. Comput. Appl. Eng. Educ. 21, 360–371. doi:10.1002/cae.20480
Heradio, R., de la Torre, L., and Dormido, S. (2016). Virtual and Remote Labs in Control Education: A Survey. Annu. Rev. Control 42, 1–10. doi:10.1016/j.arcontrol.2016.08.001
Khan (2022). Khan Academy. Available at: https://www.khanacademy.org/
Lynch, S., and Becerra, V. (2011). HEA Workshop and Seminar Series. York: Matlab Assessment for Final Year Modules
Murray, R. M. (2003). Control in an Information Rich World: Report of the Panel on Future Directions in Control, Dynamics and Systems (SIAM). Available at: http://www.cds.caltech.edu/murray/cdspanel.
Murray, R., Waydo, S., Cremean, L., and Mabuchi, H. (2004). A New Approach to Teaching Feedback. IEEE Control Syst. Mag. 24, 38–42. doi:10.1109/MCS.2004.1337856
Oliveira, P. B. d. M., Hedengren, J. D., and Rossiter, J. A. (2020). Introducing Digital Controllers to Undergraduate Students Using the Tclab Arduino Kit. IFAC-PapersOnLine 53, 17524–17529. doi:10.1016/j.ifacol.2020.12.2662
Oliveira, P. M., and Hedengren, J. D. (2019). “An Apmonitor Temperature Lab Pid Control Experiment for Undergraduate Students,” in 2019 24th IEEE International Conference on Emerging Technologies and Factory Automation (ETFA) (IEEE), 790. doi:10.1109/etfa.2019.8869247
Park, J., Martin, R. A., Kelly, J. D., and Hedengren, J. D. (2020). Benchmark Temperature Microcontroller for Process Dynamics and Control. Comput. Chem. Eng. 135, 106736. doi:10.1016/j.compchemeng.2020.106736
Persson, N., Andersson, T., Fattouh, A., Ekstrom, M. C., and Papadopoulos, A. V. (2021). A Comparative Analysis and Design of Controllers for Autonomous Bicycles. European Control Conference.
Quanser (2022).Quanser Experience Controls Take Home App. Available at: https://www.quanser.com/experience-controls
Rentzow, E., Muller, T., Golz, M., Ritz, S., Kurowski, M., and Jeinsch, T. (2021). Modeling and Control of a Highly Modular Underwater Vehicle with Experimental Results. Eur. Control Conf. 2021, 2245–2250. doi:10.23919/ecc54610.2021.9655100
Rossiter, A., Serbezov, A., Visioli, A., Žáková, K., and Huba, M. (2020). A Survey of International Views on a First Course in Systems and Control for Engineering Undergraduates. IFAC J. Syst. Control 13, 100092. doi:10.1016/j.ifacsc.2020.100092
Rossiter, J. A., Cassandras, C., Hespanha, J., Dormido, S., de la Torre, L., Ranade, G., et al. (2023). Control Education for Societal-Scale Challenges: a Community Roadmap. Annu. Rev. Control. to be submitted to.
Rossiter, J. A., Giaouris, D., Mitchell, R. J., and McKenna, P. (2008). Typical Control Curricula and Using Software for Teaching/assessment: a uk Perspective. IFAC Proc. Vol. 41, 10331–10336. doi:10.3182/20080706-5-kr-1001.01748
Rossiter, J. A., Hedengren, J., and Serbezov, A. (2021). Technical Committee on Control Education: A First Course in Systems and Control Engineering [Technical Activities]. IEEE Control Syst. 41, 20–23. doi:10.1109/mcs.2020.3033106
Rossiter, J. A., Pasik-Duncan, B., Dormido, S., Vlacic, L., Jones, B., and Murray, R. (2018). A Survey of Good Practice in Control Education. Eur. J. Eng. Educ. 43, 801–823. doi:10.1080/03043797.2018.1428530
Rossiter, J. A., Pope, S. A., Jones, B. L., and Hedengren, J. D. (2019). Evaluation and Demonstration of Take Home Laboratory Kit. IFAC-PapersOnLine 52, 56–61. doi:10.1016/j.ifacol.2019.08.124
Rossiter, J. A. (2017). Using Interactive Tools to Create an Enthusiasm for Control in Aerospace and Chemical Engineers. IFAC-PapersOnLine 50, 9120–9125. doi:10.1016/j.ifacol.2017.08.1713
Rossiter, J. A. (2011). Which Technology Can Really Enhance Learning within Engineering? Int. J. Electr. Eng. Educ. 48, 231–244. doi:10.7227/ijeee.48.3.2
Rossiter, J. (2022a). Modelling, Dynamics and Control. Available at: https://sites.google.com/sheffield.ac.uk/controleducation/
Rossiter, J. (2022b). Take Home Laboratories Enhancing a Threshold Approach to Assessment. IFAC Symposium Adv. Control Educ.
Satué, M. G., Castaño, F., Ortega, M. G., and Rubio, F. R. (2021). Comparison of Control Strategies for Hcpv Sun Tracking. Eur. J. Control 62, 165–170. doi:10.1016/j.ejcon.2021.06.008
Serbezov, A., Zakova, K., Visioli, A., Rossiter, J. A., Douglas, B., and Hedengren, J. (2022). Open Access Resources to Support the First Course in Feedback, Dynamics and Control. IFAC Symposium Adv. Control Educ.
Taylor, B., Eastwood, P., and Jones, B. L. (2013). Development of a Low-Cost, Portable Hardware Platform for Teaching Control and Systems Theory. IFAC Proc. 46, 208–213. doi:10.3182/20130828-3-uk-2039.00050
Yerolla, R., and Besta, C. S. (2021). Development of Tuning Free SISO PID Controllers for First Order Plus Time Delay (FOPTD) and First Order Lag Plus Integral Plus Time Delay Model (FOLIPD) Systems Based on Partial Model Matching and Experimental Verification. Results Control Optim. 5, 100070. doi:10.1016/j.rico.2021.100070
Keywords: education, control engineering, future trends, community vision, learning and teaching
Citation: Rossiter JA (2022) Future Trends for a First Course in Control Engineering. Front. Control. Eng. 3:956665. doi: 10.3389/fcteg.2022.956665
Received: 30 May 2022; Accepted: 22 June 2022;
Published: 11 July 2022.
Edited by:
Sebastián Dormido, National University of Distance Education (UNED), SpainReviewed by:
José Luis Guzmán, University of Almeria, SpainDavid Muñoz De La Peña, Sevilla University, Spain
Copyright © 2022 Rossiter. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: John Anthony Rossiter, ai5hLnJvc3NpdGVyQHNoZWZmaWVsZC5hYy51aw==