Hardware Reference
In-Depth Information
phases. During these phases, validation activities are being handled by involving ex-
ternal stakeholders. The participation of the external stakeholders requires synchro-
nizing agendas of all involved parties. The external stakeholders' participation is
mainly required during the
Analysis
and
Resolution
tasks. The duration of a task is
calculated based on the start and end events in the database, hence it also includes the
waiting time. The waiting time is then reflected in the increased duration of the tasks.
The throughput time drops during the
Implementation, Component testing
and
Inte-
gration testing
during which the developers perform verification activities without
contribution of the external stakeholders.
5.4 Duration in Other Projects
Besides the project P, we analyzed three other projects: P1 - P3. Each of them had
more than 1000 cases. All of the projects P1 - P3 showed similar trend in task dura-
tion and the throughput time with respect to the verification and validation activities
as observed in the project P.
6 Conclusions
In our case study, we showed that it is possible to use process mining techniques to
get more insight into a selected software development process. We compared the du-
ration of tasks and the total throughput time during different lifecycle phases. The
results showed that the duration of the validation tasks involving external stakeholders
are longer than the verification tasks performed without the external involvement.
Possible implications for the software process improvement might be that meetings
with the external partners are plan ahead, maybe on the regular basis. The problems
with synchronizing different agendas are minimized and the project progresses more
smoothly. Project issues are then solved more promptly without any extensive wait-
ing times.
Although, we expected the number of cases that are not handled according to the
documented process increases towards the end of the project, the results of our analy-
sis do not prove that. The number of cases that do not comply with the documented
model is overall high. This indicates that skipping the
Analysis
task is a structural
problem. A decrease is observed during the System testing, possibly due to the fact
that more attention to the handling of defects is given during this phase.
In the future projects, we have the intention to use process mining techniques to
analyze other important aspects of software development processes.
References
1.
CMMI Product Team. CMMI for Development, Version 1.2. CMU/SEI-2006-TR-008
(2006)
2.
Cook, J.E., Wolf, A.L.: Discovering Models of Software Processes from Event-Based Data.
ACM Transactions on Software Engineering and Methodology 7(3), 215-249 (1998)
