A survey of good practice in control education
Author
Rossiter, JA, Pasik-Duncan, B, Dormido, Sebastian, Vlacic, Ljubo, Jones, Bryn, Murray,
Richard
Published
2018
Journal Title
European Journal of Engineering Education
Version
Accepted Manuscript (AM)
DOI
https://doi.org/10.1080/03043797.2018.1428530
Copyright Statement
© 2018 Taylor & Francis. This is an Accepted Manuscript of an article published by Taylor &
Francis in European Journal of Engineering Education on 30 Jan 2018, available online: https://
doi.org/10.1080/03043797.2018.1428530
Downloaded from
http://hdl.handle.net/10072/383681
Griffith Research Online
https://research-repository.griffith.edu.au
November 21, 2017 European Journal of Engineering Education ejc˙controleducation˙aug2017
To appear in the European Journal of Engineering Education
Vol. 00, No. 00, Month 20XX, 1–22
A Survey of Good Practice in Control Education
J.A.Rossiter
a ∗
, B.Pasik-Duncan
b
, Sebastian Dormido
c
, Ljubo Vlacic
d
, Bryn Jones
a
and
Richard Murray
e
a
Dept. ACSE, University of Sheffield, UK. Email:
j.a.rossiter@sheffield.ac.uk,b.l.jones@sheffield.ac.uk,
b
Department of Mathematics, University of
Kansas, USA, Email:bozenna@ku.edu,
c
Escuela Tecnica Su perior de Ingeniera Informatica,
UNED, Madrid, Spain. Email: sdormi do@ dia. u ned.es,
d
School of Engineering, Griffiths
University, Melbourne, Australia,
e
Control & Dynamical Systems and Bioengineering, Caltech,
USA. Email: murray@cds.caltech.edu
(Received 00 Month 20XX; final version received 00 Month 20XX)
This paper gives a focu ssed sum ma ry of good practice taken primarily from engineers who are
responsible for teaching topics related to systems and control. Thi s engineering specialisation
allows the paper to give some degree of focus in the discussions around laboratories, software
and assessment, although n a t u ral ly many of the con clu si o n s are generic. A key intention is
to provide a summ a ry document or survey paper which can be used by acad em ic s as a start
point in studies of what is effective in the discipline. It is also hoped that such a summary
will will be useful to engineerin g institutions in drawing together and disseminating open
access resources that are freely available to the community at large.
Keywords: Control education, remote and virtua l laboratories, laboratory assessment ,
open access resources.
1. Introduction
Contr ol systems technology conti nues to change rapidly, causing a need to continually
revise the way we educate students. Moreover, al ongsi d e technological advances, there
has been an increasing awareness of wh at constitutes an effective learning environment
and the general perception is that this is vastly different from traditional model s. A
recognition of the importance of education is ev i d enced by the existence of educational
committees within many engineering institut i ons, including IEEE. This article focuses
on the work of several cont rol system educators who, under the umbrella of their affil-
iations with IFAC (International Federation of Automatic Control), AACC (American
Automatic Control Council), and IEEE, have shared their ideas on, and experience from,
advancin g the practice of control systems teaching and learning.
The community benefits from a consolidation of good practice in control engineering
resources in general, with education being an essential component of that. Indeed, a pre-
liminary website devoted to control resour ces was authorized and created by IFAC in
2008 (Perez, Dormido and Vlacic 2011) and updat ed recently (IFAC 2016). The IFAC
educational committees felt that a publication summarizing good practice and oppor-
tunities within control education would be useful to the community. A short summary
∗
Corresponding author.
1
November 21, 2017 European Journal of Engineering Education ejc˙controleducation˙aug2017
was presented at the IFAC world congress in 2014 ( Rossi t er , Dormido, Vlacic, Jones and
Murray 2014) and this constitutes the longer and more complete version of that work.
This article is intended to be useful primarily to those involved in deli vering education,
who may wish to updat e or improve their current practices, but it will also be of interest
to many others who may benefit from an awareness of how first degrees are delivered. A
wide range of scenarios and concepts will be covered, with an underlying focus on the
following questions:
(1) What is accep t ed good practice?
(2) What is t h e evidence and context?
(3) How and why should I try this?
1.1. Historical Context and Future Opportunities
In the past, there were relatively limited opportunities and mechanisms for educational
delivery and assessment. Recently, there has been significant pressure to understand the
learning process, app l y research t o it, and exploit advances in technology. Consequently,
there have bee n substantial advances in the learning environment and the awareness of
academic staff of how to support student learning. It is now commonplace for new aca-
demics to achieve formal qu ali fi cat i on in learning and teaching (in the UK, for example,
this is ratified by the higher education academy www.heacademy.ac.uk). This encourages
staff to reflect in detail on learning outcomes, by considering questions such as: (i) what
do I want the students to be able to do, and why? (ii) how can I be sure they achieve
this? (iii) what effective good practice is t her e? Staff have both i n centive and su p port to
pursue good practi ce. One core example of the increased focus on pedagogy is the discus-
sions on the role of a conventi on al didactic lecture (Crouch and Mazur 2001; Lancaster
and Read 2013); traditional modes of delivery still have a role, but a r ed uced one.
In parallel with pressures for academics to become ‘professional’ educators, as well as
research er s, there have been significant changes in the Learning & Teaching technology
domain. Leaving aside smaller issu es such as the ease with which staff can now replicate
notes for distribution, the power of computing has had the most impact. Since the desktop
computer became affordable ar ou nd the early 1990s, there has b een the potential for
universities to allow a massive increase in student access to computing. One simple benefit
is the opport un i ty to include more complex and interesting prob lems on written exams,
by allowing access to su it ab l e software during assessment to aid with tedious number
crunching. There is evidence that the community at large is now begin n in g t o r eali ze
this on a broader scale and thus the integration of software packages into assessment s
and even formal examinations is becoming the norm as opposed to a rarity (Lynch and
Becerra 2011; Rossit er, Giarouris, Mitchell and Mckenna 2008).
Similar opportunities exist with regard s to laboratory access. While time spent sitting
in front of equipment will always be limited at most universities due to a combination
of financial, sp ace, and timetable restrictions, there is now the potential to allow remote
access to ‘laboratory-like’ activi t i es 24/7. Indeed, the increasing power and speed of t h e
Inter net allows for real-time access to data, laboratories (Gustavsson 2009), and related
resources. Software has also improved so that the average academic can author relat ively
advanced simulators (Cameron 2009; Guzmand 2006; Khan and Vlacic 2006; Goodwin
2010) quickly and without need of coding experti se. This has led to many innovati on s
such as remote access laboratories, virtual laboratories, interactive online laboratories,
computer-aided assessment, online audio, video, and text files, and so forth (de Jong,
Linn an d Zacharia 2013; Mathtutor 2012; Rossiter 2016; Murray 2013). Despi t e the
wealt h of available suggestions in t h e literature about the imp lementation of all this new
tech nol ogy, many institutions have yet to really avail themselves of it.
2
November 21, 2017 European Journal of Engineering Education ejc˙controleducation˙aug2017
The rapid advances in computing technology also mean that departments now have the
potential to develop very cheap equi p ment (Hi l l 2015; Reck and Sreenivas 2015; Egerstedt
2014). No longer does each experiment need an expensi ve IO card or similar; instead one
can plug into the computer directly via a USB port. This allows for the production
and distribution of mul ti p l e copies of real equipment, which can even be taken home by
students (Taylor, Jones and Eastwood 2013; Eichlere 2013).
This article also has a brief section that aims to recognize the incr easi ngl y multi-
disciplinary skill requirements for modern engi neer s. In the same vein, systems and con-
trol skills are r equ i re d in many disciplines outside of engineering. Hence, there are oppor-
tunities for engineers to recognize how their skill set s can be more broadly applied and,
conversely, opportunities to introduce novel ‘systems’ modules into other areas (Murray
2013).
1.2. Summary of Contr i buti on
This arti cl e summarizes best practice in tertiary education with a specific focus on the
delivery of systems and control topics. The first section is generic (that is, applicable
to all disciplines), but the authors feel it will b e useful to readers who may not come
across such concepts otherwise. It focuses briefly on the largest student-staff interaction,
which is through the lect u re (Crouch and Mazur 2001); this is where we have the most
time and opportu ni ty to influence students. It also covers online learnin g to ol s and the
use of software in assessment. Some examples from within control t eaching are given.
The next section considers st u dent engagement and the use of technology in helping
students learn better, within a systems and control scenario. The following two sections
focus on laboratories and equipment to support the learning of systems and control
and, in particular, how modern technology prov i des opportunities for cheap er and more
accessible activities. There is a dedicated discussion of ‘take-home-laboratories;’ that is,
hardware students can have access to 24/7. The final section touches on the evangelizat i on
of systems and control to disciplines outside of engineering.
This article will not discu ss issues linked to accreditation (UK-SPEC 2016; ABET
2016(@; ENAEE 2016), industrial involvement and overall curriculum design/holistic
student development.
2. The Design and Supply of Staff Involvement in Module Teaching
This section focuses on generi c principles related to education, beginning with what
constitutes an effective l ect ur e, particularly in the context of something like control. A
critical point is that diversity is usually good and thus lecturers are encouraged not to
adopt a single technique; often, a variety of approaches within the same course is best .
There is also a b r ief discussion of virtu al learning environments (VLE), the opportuni-
ties these offer to improve the student learning experience, and the u se of software in
assessment.
2.1. Didactic Lectures
Employers are interested in students’ ability to abstract, see the bi g picture, analyze real
world problems, learn independently, account for issues such as risk and reliability, etc.
(Pan el 2013). A poorly delivered didactic lecture can, conversely, give the impression that
the lecture content is the totality of what students need to know and thereby encourage
a memorize-and-regurgitat e approach to learning (Rossiter and Gray 2010). In a similar
3
November 21, 2017 European Journal of Engineering Education ejc˙controleducation˙aug2017
vein , poorly designed tutorial questions may not test problem solving but rather students’
ability to r epl i cat e/copy from notes (Kawski 2013).
In a feedback loop, a core component is the control law; ul t imat el y, it is this that gov-
erns s y st em behavior and not the measurement/information. Within a learning scenario,
the lecturer provides the information or measurement but it is the student who acts as
the control law. Thus, effective learning can only take place when the student is act ive! A
passive student is effectively open-loop an d may learn very little. It is recogn ized that stu-
dent engagement drops rapidly aft er 15 minutes (Huxham 2005) of di d act ic presentation.
Hence, the overall portfolio of lectures must not be overloaded with didactic presentation
tech ni q ues that convince students, by examp le, that passivity and replication is what is
valued. For those who are interested, there is a chapter in (Abdulwahed 2010) on links
between learning and feedback loops.
Despite these concerns, didactic presentation modes still play an important role and
should not b e removed ent i r el y. After all, an effective feedback loop does benefit from
good quality measurement/information. Important considerations in the delivery of di-
dactic lectures include the following.
(1) Many stu dents do not parse formulae correctly; what they ‘say’ does not match
what they write. Watchi n g a lecturer carefully write/talk through a solution on a
black board with the correct language will help them learn t he correct associati ons
and approaches.
(2) A screen dump of a solu t ion consisting of several steps can confuse or switch students
off. By writing a solut ion out during the lecture, an instructor is forced to go at a
pace where students can clearly identify the steps and thinking and thus foll ow the
argument. This hel p s students become active and not passive!
(3) There are some important messages that the lectu rer wants to be sure ar e presented
correctly; ensure no measurement error!
(4) Good pract i ce suggests lecturers should face students and thus use an overhead
projector, pen-enabled laptop (Wilson and Maclaren 2013), or smart screen. Agai n,
this encourages engagement and reduces passivity.
2.2. Online Lectures or No Lectures and Sharing Learning Resources
One of the most obvious weaknesses of lecture deliveries is that they happen onc e, at a
fixed time, and if you miss them there are no second chances. Even for those present,
a temporary lapse in concentration can mean the failure to note an impor t ant point.
Moreover, difficulty picking up the nuances of the lecturer’s speech can result in a mis-
understanding or gap in the notes. These problems are known to be amplified in the case
of international students (Rossiter 2009) who are less flu ent in the language of instruc-
tion. In such cases, there is substantial evidence that the recording of lectures is hugely
valuabl e (Fidler, Middelton and Nortcliffe 2006; Middleton 2013) as students can listen
to the lecture again later and correct any initial misunderstandings, reinforce the sounds
and interpretations of keywords, update and correct their notes, and so on. This also
helps those who have no language difficulti es.
Staff may begin to wonder whether recording lectures el i mi nat es the need for a face-to-
face lecture altogether. There is evidence (Parson 2009) that lecture recording does not, in
fact, reduce class attendance and indeed this matches the authors’ personal experiences.
Recordings are best us ed as a complement to, rather than a replacement of, in-person
attendance of lectures. Nevertheless, there remains a question about the precise role
of recordings within an overall curriculum delivery and this is of di r ect interest to the
contr ol community. To what extent would a validated repository of learn i ng resources
in systems and control be useful to both academic staff and students (Saund ers and
4
