1.      Introduction

Pursuing a PhD can feel like walking on a tightrope:
exhilaration and adrenaline arising from the challenge can drive one forward,
but the fear of failure and of a ruinous fall might be paralyzing [1].

The psychological distress faced by graduate students
has multiple roots, such as difficulties in maintaining a work-family balance,
lack of inspirational guidance, uncertainties about future career outcomes,
etc. [2]. Addressing, not to mention solving, these challenges cannot be done
with a one-size-fits-all strategy; each PhD is a unique experience by a unique
individual [3].

However, one common feature is shared by all research
projects, namely the strive to push boundaries of science forward. This is a
highly uncertain and sometimes unpredictable task and for this reason
traditional management strategies, such as ‘waterfall’ methods [4] and Gantt
charts [5], tend to fail. PhD projects do not follow the plans and so far there
is no clear risk-mitigation system to support the graduate student when this
happens.

In the present
work an Agile approach to graduate education, in general, and the
management of PhD projects, more specifically, is proposed. The goal is to draw
the attention of an Academic audience to this management philosophy, as a
promising instrument for potentially enhancing the productivity of PhD
projects, while at the same time improving the psychosocial wellbeing of
graduate candidates.

2.      Agile principles

When the Agile Manifesto [6] was first published, 18 years
ago, it started a disruptive revolution in the world of software development.
One of the core principles at the heart of this philosophy is ‘Responding to
change over following a plan’. Having an Agile mindset means being
flexible, responsive, adaptable. In an Agile world, progress is measured
through the attainment of concrete results and working outcomes, rather than by
ticking off successive tasks on a to-do list. An early, limited result, which
can be swiftly elaborated in a later stage, is more significant than a perfect
result achieved only by the deadline of the project [7]: the first allows for
feedback during the project, while the latter simply does not. It is tempting
to transpose the characteristics of Agile management as described above
to graduate research projects.

Based on the available literature [8], some principles are herein
proposed to translate the concepts of Agile into practical guidelines
for the daily activities of PhD researchers. The suggested steps to be
implemented are the following:

1.              
A big block of work is divided into several
‘layers’ of activities. What defines a layer is that it ends with the
attainment of a concrete result, as opposed to the traditional waterfall
methodology where activities are organized purely based on the timeline. Each
layer will be undertaken in a specific, limited period of time (e.g. 2-12
weeks), named a sprint.

2.              
Sprint planning: the
PhD student, the supervisor and any other stakeholders (for example a postdoc,
industrial partner, or masters student) meet for around 30 minutes aiming to set
the goal of the sprint and its duration. Everybody must  agree on these
two points, to ensure that the entire research team is on the same page and
that prospects are the same for everyone. As part of the sprint planning,
the sprint review meeting (see step 5) can already be scheduled.

3.              
Sprint execution:
work! Maximum focus is dedicated on each particular task for a definite amount
of time.

4.              
Weekly ‘scrum’:
meeting between the PhD student and the supervisor for a maximum of 15 minutes,
but ideally every week to keep the momentum, (e.g. the same timeslot every
week, outside of ‘conventional’ working hours to make sure there are no
meetings, teaching activities etc. which may get in the way). The point of this
meeting is to be short and efficient: better avoiding chairs so that everyone
stands up and there is no risk of getting too comfortable and overextending the
meeting. No laptops, not even papers. Only three questions need to be answered:

a.     
What was done the previous week in order to
contribute to the goal?

b.     
What will be done next week in order to
contribute to the goal?

c.     
Are there any impediments?

5.              
Sprint review + sprint
retrospective + sprint planning: at the end of the sprint, a
meeting is planned including the PhD student and  all the stakeholders involved
to discuss the obtained outcomes and whether those are in line with the goals
and expectations set up in the previous planning (review). This is the
moment to go into the specifics and brainstorm altogether. It is crucial to
discuss the encountered difficulties, in order to make sure the next sprint
is continuously improved compared to the previous one (retrospective). This
is the phase when impediment removal (= problem solving, in Agile jargon)
is dealt with: sincerity and transparency are key. As previously mentioned, Agile
is all about adapting to change, and plans can change. Go back to point 1
and restart the planning, addressing the next layer of work in a new sprint.

The conceptual
difference between a traditional and an Agile approach to PhD management
is summarized in Figure 1.

Figure 1. Conceptual difference between traditional
PhD management (red) and Agile PhD management (green).

3.      Agile PhD case studies

To illustrate
our methodology, we provide the application of the Agile principles to
three main activities of a researcher (for example in chemical engineering),
see Table 1. These case studies
offer just a simplified view of some activities which can be encountered in a
research project, but are meant to exemplify how the Agile approach aims
at finalized, even if preliminary, results in the early stages of work.

Table 1.
Comparison of traditional and Agile approach for three recurring case studies
in graduate education of, among others, chemical engineers.

Currently
the Agile principles are in use in the group of Catalytic Reaction
Engineering (CaRE) of Ghent University, with the aim of evaluating the
long-term potential benefits of this approach. Some field-specific challenges
are being tackled such as top-down construction of kinetic models for
heterogeneously catalyzed reactions or the inclusion of particle-scale
transport phenomena into simulation codes for multiphase reactors. The results
obtained so far have been evaluated only qualitatively. However, the
participants have reported an increase in productivity and, thanks to the
enhanced number of contact moments, an improved perception of the working
environment, both of which are promising and encourage the establishment of
collaborations with other Academic institutions. These collaborations aim at quantitatively
evaluating an Agile management of research projects on a statistically
relevant number of subjects.

4.     
Conclusions

The application of an Agile management style
to research projects could positively impact the Academic environment in terms
of increased flexibility and adaptation to the ever-changing challenges to be
tackled in science. Particularly, graduate students and their supervisors are
encouraged to follow this approach as risk-mitigation plan to face the
endeavors of PhD projects.

Acknowledgments

The authors would like to thank Luca Comegna and
Daria Otyuskaya for the useful discussions, and the members of the CaRE group
at the Laboratory for Chemical Technology (Ghent University) for being the
first ‘guinea pigs’ of the proposed approach.

References

[1] C. Woolston, Graduate Survey: a Love-Hurt Relationship, Nature
(2017)

[2] K. Levecque, F. Anseel, A. De Beuckelaer, J. Van der Heyden, L.
Gisle, Work Organization and Mental Health Problems in PhD Students, Research
Policy (2017)

[3] L. Constantin, How to handle the one-size-fits-all PhD, Nature
Career Column (2018)

[4] C.D. Tupper, Data Organization Practices,  Data Architecture:
from Zen to Reality, Morgan Kaufmann (2011)

[5] H.L. Gantt, Work, wages and profit, Engineering Magazine (1910)

[6] Manifesto for Agile Software Development (2001)

[7] J. Sutherland, SCRUM The Art of Doing Twice the Work in Half the
Time, Random House Business Books (2015)

[8] Agile Project Management:
Quick Start Guide, Albany (2016)