Professional Documents
Culture Documents
May 2015
Page 1
Table of Contents
• Executive Summary
• About The Study
• Participant Demographics and Context of Agile Application in Practice
• Scope and Phase of Agile Implementation
• Agile – What and Why?
• Agile Methods and Practices
• Tools used
• Experiences and Lessons Learned
• Comparison with the previous year’s results
• Summary and Conclusion
• Annex, Acknowledgements, and Disclaimer
Agile Implementation
• Agile implementation has gained a foothold at all development organization levels
• Agile is typically first introduced in an internal project (e.g. platform development) and then within a customer project
• Experience with the use of Agile methods and practices ranges from 6 months to 10 years
• Agile is applied in all domains, not only in Multimedia series development, but in most cases applied only to software development
• Agile seems to be used mainly at the beginning of series development (A and B Sample); in later phases (C and D sample), more traditional
development methods are equally applied
• Unanimous view of the respondents: Following Agile methods and principles “blindly” as described “by the book” would have led to failure;
therefore, respondents do “cherry picking” and implement agile practices which are useful for them
• Major concerns regarding Agile: Inability to scale and lack of up-front planning
Next steps
• Next steps in the automotive world are to offer solutions to the scalability concern and improving in the field of automated testing
Page 4
About The Study
Page 6
About The Study (2)
Survey Process
Selection of approx. 60 Interviews and
key representatives Online surveys
Nov 2014 –
Feb 2015
Analysis of the data
Study Report collected. Conclusions
April 2015 sent to participants
in a detailed report
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
• The main reason for adopting Agile is that • On average, organization apply Agile for three
traditional software development methods are years; however, experience with the use of
not flexible enough to address the emerging Agile methods and practices ranges from 6
challenges in software development, e.g. months to 10 years
increasing complexity, constantly changing
requirements, etc.
Integrated Systems
Body Electronics
Application Types
Multimedia
and Services
10% • Driving Assistance/Automatic Driving
• Intelligent Mirror
5% • Active Safety
0%
80%
79%
74%
70%
62%
60%
52% 52%
50%
40%
Pre-series Dev.
30%
D Sample
A Sample
C Sample
B Sample
20%
10%
0%
38%
45%
55%
62%
AnAnaverage
averageteam
teamdoing
doing Agile
Agile is Developers are typically assigned
locatedat
located in one/multiple
one/distributed over
site(s) to one/multiple team(s)
multiple locations
29%
40%
60%
71%
• A median sized Scrum team consists • A median sized Kanban team consists
of 8 people of 8 people
• The largest stated Scrum team has • The largest stated Kanban team has
about 20 members, the smallest only about 40 members, the smallest only
3 members 3 members
Team(s) level
Whole 30%
Development 38%
Organization
33%
Project(s) level
SB 7%
Abbreviations:
Kanban 55% • TDD = Test Driven Development
• FDD = Feature Driven
Scrum 79% Development
• XP = Extreme Programming
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
• SB = Scrumban
Page 26 Agile in Automotive – State of Practice 2015, Kugler Maag Cie
Agile Methods and Practices (5)
Scrum is nearly always tailored; whole scope of Kanban is increasingly being used
Scrum Kanban
• Scrum is rarely used “by the book” (Scrum • Kanban is mainly used for a continuous work
Guide); if at all, then it is used on team level; flow, e.g. Maintenance and Support (Bugs,
processes and roles are more or less tailored, Change Requests)
depending on the project environment; usually
a sub-set of agile ceremonies are in use • Kanban is sometimes mentioned only as a
visualization mechanism
• Scrum is often used in combination with other (Board); however, the real scope of Kanban
Agile practices Continuous Integration, Test goes beyond pure visualization
Driven Development, etc.
• Compared to last year’s study, more and more
“full application of Kanban principles”, i.e. limit
WIP, etc., are mentioned
Iteration/Sprint Pre-Planning
60%
Iteration/Sprint Planning
50%
Iteration/Sprint Review
40%
Daily Stand-up
30%
Retrospective
20%
10%
0%
Velocity 76%
Release Burndown Chart 52%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
50% 21%
Product Owner
Scrum Master
20%
• Single Product Owner versus hierarchy of • Scrum Master is sometimes Software Project
Product Owners (led by Leader, Team Member not part of the
Product Manager=Chief Product Owner), or development team
Product Owner managing a “stakeholders”
group
• Product Owner is sometimes Software Project • It is difficult to assess whether the standard
Leader, Group Leader, roles (e.g. Project Manager) mapped to agile
Senior Management roles (Scrum Master/Product Owner) are really
working according to Scrum rules
• Only in very rare cases is the QA role part QA tasks are taken into account in
of the agile team, with all tasks taken into the sprint and included in planning
account during the sprint but still having an
independent reporting channel. In most Independent Product QA group owns
cases, QA is independent and organized validation responsibilities. It does not
traditionally follow an agile methodology
• ISO 26262 is not perceived as being difficult So far it is assumed that it can be
to integrate with Agile. done without compromising on
Functional Safety
• For most interviewees ISO 26262 is not in
contradiction with Agile, or it was not Our AUTOSAR project was developed as
applicable for them. All levels of ASIL were ASIL D, was also Agile and worked well.
mentioned to be within the scope
PTC Integrity
30%
20% 17%
12%
MS Project
10%
Others
0%
30% 29%
IBM Synergy
25%
Clear Case
20%
17%
15%
12%
PTC Integrity
10%
Subversion
7%
Others
5%
Git
0%
50%
50%
43%
40%
30%
Gmake
Maven
21% 21%
20%
Visual Studio
10%
10% 7%
Jenkins
Others
Cmake
0%
70% 67%
60%
50%
40%
Bamboo
30%
20% 17%
Jenkins
Others
10%
5%
0%
40%
Quality Center
33%
PTC Integrity
30%
20%
14%
10%
10% 7%
Jenkins
Others
5%
Xunit
0%
• The majority of respondents mentioned that Getting the tools right for the first
following Agile methods/principles “blindly” as pilot was hard. The first pilot failed
described “by the book” (Scrum Guide) would have because of that. Once the tools were
led to failure right, adoption went through the roof.
• The main perceived “risk” areas for failure are lack Some teams did not accept Agile -
of understanding what Agile means, and because of the culture of the team
appropriate “soft” skills leader, not because of the team.
Team leader was afraid to provide
visibility about what was going on…
• The main pointed-out barriers were trying to fit
agile elements into a non-agile environment, ability
Difficulty to build a team-based
to change organizational culture, project complexity
culture. Old management
and customer collaboration
behavior still in place
• All respondents confirmed the continuation of The only way to get a working tested
their agile activities; e.g. pilot evaluation, next software in three weeks is to have
pilot, continuous roll-out, or continuous automated test. Goal: 90% of automation.
improvements of the current processes.
Stabilization of the process, expansion to
• Scaling and Automated Testing is getting more other locations, multiple Scrum teams
and more a crucial topic
Adopt additional agile principles, put
more discipline into scrum, launch a
metrics program, adopt better tools
SB
XP
6% 2015
7%
6%
SB 7%
Kanban 39%
Kanban
55%
Scrum
Scrum
79%
95%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
70% 67%
2015
61%
60% 56%
50%
44%
40%
33% 33% 33%
31%
Integration Test
System Design
System Test
Integr. Test
20%
Software
Software
Software
Software
Analysis
Design
10%
Test
0%
XP
29%
33%
SB 43%
2014 56%
2015
Kanban
11%
29%
Scrum
Stabilization Roll-out
Page 60 Agile in Automotive – State of Practice 2015, Kugler Maag Cie
Comparison with the previous year’s results (5)
JIRA and Jenkins are the big winners, significant tool change in Test Automation
Comparison of the top Tools
80%
2014
70%
Continuous
Feature Driven Integration 67%
Development
60%Test Driven
Development52%
2015
XP
50%Feature Driven
50%
Development
40%
SB
XP
Project/CR/PR/Task Mgmt.
33% 33%
28%
Test/Test Automation
Config. Management
30%
SB
20%
Kanban 17%
Kanban
Subversion
Subversion
Integration
Own Tool
Scrum
Jenkins
Jenkins
Jenkins
Jenkins
Jenkins
10%
Scrum
JIRA
JIRA
Build
0%
• Kanban is mainly used where there is a • In most cases, the existing roles are kept,
continuous flow of work, e.g. maintenance and Product Owner/Scrum Master scope of work is
support (bugs, change requests), where work in added to the existing roles, and it seems to
fixed iterations does not make sense or work well. Product Owner and Scrum Master
resources cannot be committed roles are rarely defined in the same way as they
are described in the Scrum Guide
• Agile is applied in all domains, not only in
Multimedia series development; however, Agile • Typically, the Product Owner role is handled by
seems to be used mainly at the beginning of the Software Project Leader and Scrum Master
series development (A Sample and B Sample) role either by the Team Leader or a completely
new person. Social skills of Scrum Masters are
• Agile is typically first introduced in an internal perceived as essentials for success
project (e.g. platform development) and then
within a customer project; there are no major • The most common team setting is that one
organizational changes needed to run Agile team is working on one product and supporting
projects – apart from roles one or more projects. Agile teams are typically
spread over multiple sites and developers are
assigned to only one team. On average, Kanban
teams are 25% larger than Scrum teams
• Co-location is not a must, the crucial aspect for • Almost all Agile attempts are successful;
building teams is the need for communication management commitment and openness to
among its members change are key factors; a majority of
respondents mentioned that following Agile
• Tools and Infrastructure are a key precondition methods / principles “blindly” as described “by
for successful Agile implementation and are the book” (Scrum Guide) would have led to
often on the top of the agenda of an Agile failure. Major concerns regarding Agile: Inability
transformation programme to scale and lack of up-front planning
• A big variety of tools can be found in Agile • The scope of the application of Agile practices
projects. Though it is difficult to identify the and methods in the automotive industry is
leaders for some areas, the management tool mainly limited to software development
Jira seems to be the most popular one; for
integration it is Jenkins • Full benefits of Agile may be achieved when
applying it beyond software development and
• Automotive SPICE as well as Functional when applying it as a OEM/Supplier co-
Safety/ISO 26262 are not perceived as difficult operation
to combine with Agile
• Next steps in the automotive world: Scaling and
Automated Testing
Page 65 Agile in Automotive – State of Practice 2015, Kugler Maag Cie
Summary and Conclusion (4)
An alternative summary
Agile Manifesto/Principles and Automotive Expectations and Benefits
Reading the Agile Manifesto with Automotive eyes can create some The expectations regarding Agile are very ambivalent. On the one
controversy. Some principles are viewed quite negatively in terms of hand it is seen as the silver bullet that will solve all our software
applicability to Automotive, specially when misread or misunderstood. problems while, on the other hand, there is a very strong negative a-
The most consensual principles are iterative and incremental priori attitude that it is not applicable to Automotive. Each individual
development with shorter feedback cycles, a team-centric approach (developers, managers) has their own expectations based on a
(empowerment, self-organization, learning, communication, inter- personal interpretation of the principles. Productivity, Quality and
actions, etc.), building quality from the start, and continuous Transparency are the most cited expected improvements. However,
improvement. the improvements mostly mentioned are related to human aspects;
Agile is above all a mindset that is supported by principles and i.e., staff morale, collaboration, etc. Agile puts the developers and
methods. Moving away from the traditional development approach is teams back in the forefront. Few organizations can prove
commonly viewed as a “MUST” to manage the rising complexity and quantitative improvements.
constantly growing pace of changes. Agility is a competitive weapon for
automotive companies; however, the principles from the Manifesto
and methods need to be tailored to the constraints (SOP, complex Experiences and Lessons Learned
supply chain, etc.). Experiences with Agile vary from six months to up to ten years. The
early adopters built up new organizations almost from scratch,
taking into consideration current trends, including Agile (those
Agile in Automotive: Scope and Methods showing high productivity improvement). The “early majority” is
The scope of application of Agile in Automotive is mainly software. This piloting Agile in a an incremental way.
is rather legitimate since methods like Scrum are coming from the Is Agile the “Flavor of the month”? Well, this will strongly depend on
software world. The extension to other disciplines; e.g. System, the “change management” approach used.
Hardware, is perceived as difficult since those activities are managed An implementation of Agile addressing, in sequence, first values,
sequentially and the overall automotive ecosystem is rather traditional, then principles, and finally methods and practices, will be prone to
not to say conservative. success. Other approach are just “methods patches” and will not
All existing methods or frameworks; e.g. Scrum, Kanban, etc. have been sustain long-term.
adapted to automotive constraints. The choice of suitable practices is
cherry-picking based, applying what fits best to a given environment.
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Page 68 Yes
Agile in Automotive – State of Practice 2015, Kugler Maag Cie No No Answer
Acknowledgements
Note: Small rounding errors may occur through the use of integer percentages.
Page 71
KUGLER MAAG CIE GmbH
Leibnizstr. 11,
70806 Kornwestheim, Germany
information@kuglermaag.com
www.kuglermaag.com
Page 72