IEEE Signal Processing - May 2018 - 110

SP education
Sithan Kanna, Wilhelm von Rosenberg, Valentin Goverdovsky,
Anthony G. Constantinides, and Danilo P. Mandic

Bringing Wearable Sensors into the Classroom:
A Participatory Approach

B

y bringing research into the curricu-
lum, this article explores new oppor-
tunities to refresh some classic signal
processing courses. Since 2015, we in the
Electrical and Electronic Engineering
(EEE) Department of Imperial College
London, United Kingdom, have explored
the extent to which the level of student
engagement and learning can be en-
hanced by inviting the students to per-
form signal processing exercises on their
own physiological data. More specifical-
ly, using new wearable sensor technology
and video instructions as an experiment
guide, the students are asked to record
their electrocardiograms (ECGs) and per-
form both time- and spectral-domain
estimation tasks on their own real-world
data. In this way, the students not only
gain experience with recording hardware
and sources of signal contamination
(baseline wanders and artifacts), but they
also are highly motivated by being kept
in the loop and through their part owner-
ship of their course.
In addition to bringing biopresence
into the curriculum, this approach has also
broadened the students' perspective on
research and employment opportunities
in, among others, next-generation health
care. All of the material necessary to
implement this coursework can be found
at www.commsp.ee.ic.ac.uk/~mandic/
Biopresence_Material.htm.

Motivation
The recent advances in sensor technology
and the fast-evolving way in which we
generate, process, and revise information
Digital Object Identifier 10.1109/MSP.2018.2806418
Date of publication: 26 April 2018

110

is bound to have a major impact on how
we teach and learn. Given the relative
maturity of the signal processing field
and the availability of competing multi-
media educational resources, a common
challenge in modern signal processing
education is to strike a balance among
the background physics intuition, mathe-
matical rigor in lectures, practical rele-
vance of the material, and student
engagement. This tradeoff arises from
both the diversity in student learning
styles and the requirements imposed by
the future careers of these students.
For example, M.Sc. courses typically
attract students from vastly different
backgrounds, while at Imperial College
we encourage students to take cross-
departmental courses (EEE, computing,
bioengineering, etc.); this both requires
particular care not to be EEE specific
and adds further difficulty in creating
one-size-fits-all course material. In addi-
tion, in an advanced signal processing
(ASP) course, we prefer to include ele-
ments that prepare the students for jobs
that may not even currently exist but
will be in much demand in the next four
to five years [1].
The increasing exposure to popular
technological advancements, such as big
data, the Internet of Things, and wear-
able devices, is naturally asking questions
about the relevance of some classic sub-
jects in modern real-world applications
[2]. In other words, it cannot be assumed
that students will be able to appreciate
the scope of their courses without explic-
itly building a bridge between the mate-
rial in lessons and cutting-edge research.
This all calls for more effort in bringing
research into the classroom. From our
IEEE Signal Processing Magazine

|

May 2018

|

own experience, it is biopresence that
not only has a broad spread of signal
processing demands but is most conve-
nient to implement and most enjoyed by
the students.
Recent articles on signal processing
education aim to demystify and unify
some compulsory material in signal pro-
cessing courses [3]-[5]. Our aim here is
to reflect on the opportunities provided
by the recent sensing technologies [6]
and, in doing so, meet the demands of
the new technologically orientated class-
room. Indeed, standard signal process-
ing courses were developed with radar,
array processing, and communications
problems in mind (all of which require
EEE-specific knowledge); however, to
maximize the impact of our area we
ought to facilitate the adoption of signal
processing tools by other areas, such as
bioengineering [7], [8].
Our approach focused on inviting the
students to collect and analyze their own
physiological data (ECG and respiration)
using the latest signal acquisition devices,
enabling students to
1) apply classic signal processing tools
taught in the lectures to their practical
assignments in a physically meaning-
ful and fun way
2) appreciate the relevance of their
degree with respect to technological
advancements
3) take part-ownership of their taught
courses
4) engage in curiosity-driven learning
and explore feasible extensions and
new applications of their taught mate-
rial on their own data
5) become aware of ethical issues and
the societal impact of their work.
1053-5888/18©2018IEEE
http://www.commsp.ee.ic.ac.uk/~mandic/Biopresence_Material.htm http://www.commsp.ee.ic.ac.uk/~mandic/Biopresence_Material.htm
Table of Contents for the Digital Edition of IEEE Signal Processing - May 2018
