Transparency and Contracts: Continuous Integration and Delivery in the Automotive Ecosystem
Download as PPTX, PDF1 like205 views
This document discusses the challenges and benefits of increasing transparency in inter-organizational continuous integration and delivery (CI/CD) within the automotive ecosystem. The authors address risks, information sharing needs, and the role of contracts, concluding that enhanced transparency is crucial for improving collaboration, project quality, and efficiency. The study suggests that flexible contracts and industry-wide standards may facilitate better outcomes in automotive CI/CD processes.
Transparency and Contracts: Continuous Integration and Delivery in the Automotive Ecosystem
- 1. Transparency and Contracts: Continuous Integration and Delivery in the Automotive Ecosystem Eric Knauss Jacob Juul Rogardt Heldal Patricia Lago Patrizio Pelliccione Rob van der Valk ICSE 2018, May 31, Gothenburg, Sweden Contact speaker: eric.knauss@cse.gu.se | @oerich
- 2. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 2 The Automotive Ecosystem • Currently changing • 80-90% of innovations based on electronics • Complex value-chains • Suppliers for mechanical, electrical, and SW components • In-house development • Knowledge about building a car is distributed Motivation and Goal
- 3. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 3 Development Verifica1on Feature Func1on Electrical Architecture System ComponentParallel development: suppliers + In-house development “Big-bang” integra1on! Collabora1on based on contracts Pressure to change - Shorten time-to-market - Maintain highest quality - Increase flexibility Challenge - Apply beyond scope of local teams - Apply beyond scope of individual companies Reduce cycle times and improve feedback - But: What information to share? - Accountability, openness, efficiency Transparency Motivation and Goal
- 4. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 4 • RQ1: What are risks and/or benefits of increasing transparency in an inter-organizational CI&D setting? • RQ2: (How) can sufficient information be provided in inter-organizational CI&D collaborations? • RQ3: Are contracts an impediment for scaling CI&D across company boundaries? • RQ4: Are industry-wide standards and processes shared among organizations an enabler for inter-organizational CI&D? Between partners in automotive value-chain that collaborate in continuous integration and delivery Transparency Motivation and Goal
- 5. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 5 CI/CD Support Transparency Support Often focused on small teams Fowler, Neely et al. RE differs: classic vs. oss/seco Robertson, Ruhe, Scacchi How to scale / include hw? Debiche et al. Focus: emergent developers/knowledge Haenni et al., Minto et al., Sadi et al., Treude But challenges! Eklund et al., Roberts, Rogers Less focus: Emergent reqt- contributors, across org.-boundaries, transparency Kwan et al., Knauss et al., Linåker et al., Dabbish et al., Husseini et al. Continuous delivery Humble et al., Rissanen et al. This will be important for the automotive ecosystem Damian et al., Boden et al., Kilamo et al. Related Work
- 6. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 6 Research Method Phase What it is What we did Thematising Decide purpose and subject matter In-depth knowledge about CI/D and Transparency in automotive
- 7. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 7 Research Method Phase What it is What we did Thematising Decide purpose and subject matter In-depth knowledge about CI/D and Transparency in automotive Designing Define research method in 7 steps Interview survey, transcription, analysis protocol, verification, reporting Interviewee
- 8. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 8 Research Method Phase What it is What we did Thematising Decide purpose and subject matter In-depth knowledge about CI/D and Transparency in automotive Designing Define these seven steps Interview survey, transcription, analysis protocol, verification, reporting Interviewee Interviewing Define interview style and conduct interviews Semi-structured, interview guide, select interviewees that are knowledgeable and cover a broad set of competencies
- 9. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 9 Research Method Phase What it is What we did Thematising Decide purpose and subject matter In-depth knowledge about CI/D and Transparency in automotive Designing Define research method in seven steps Interview survey, transcription, analysis protocol, verification, reporting Interviewee Interviewing Define interview style and conduct interviews Semi-structured, interview guide, select interviewees that are knowledgeable and cover a broad set of competencies Transcribin g Transcribe recordings …in English
- 10. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 10 Research Method Phase What it is What we did Thematising Decide purpose and subject matter In-depth knowledge about CI/D and Transparency in automotive Designing Define research method in seven steps Interview survey, transcription, analysis protocol, verification, reporting Interviewee Interviewing Define interview style and conduct interviews Semi-structured, interview guide, select interviewees that are knowledgeable and cover a broad set of competencies Transcribing Transcribe recordings …in English Analysing Extract knowledge from transcripts Based on Saldana, Atlas.ti, coding to describe trains of thought, use to support/negate prepositions. Prepositions defined based on experience in project consortium
- 11. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 11 Research Method Phase What it is What we did Thematising Decide purpose and subject matter In-depth knowledge about CI/D and Transparency in automotive Designing Define research method in seven steps Interview survey, transcription, analysis protocol, verification, reporting Interviewee Interviewing Define interview style and conduct interviews Semi-structured, interview guide, select interviewees that are knowledgeable and cover a broad set of competencies Transcribing Transcribe recordings …in English Analysing Extract knowledge from transcripts Based on Saldana, Atlas.ti, coding to describe trains of thought, use to support/negate prepositions. Prepositions defined based on experience in project consortium Verifying Ascert validity, reliability, generalizability Discussed in threats to validity Reporting ... …
- 12. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 12 What are risks and/or benefits of increasing transparency in an inter- organizational CI&D setting? Research Question 1 • Proposition 1: Increasing inter-organisational transparency of information is a necessary condition for inter-organisational CI&D Results (RQ1)
- 13. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem Synergy (agile collab.) Inter-org. CI&D Legend Software dependent (direct) Software driven (indirect) Software agnostic Information transparency Positive impact (dashed = indirect) Negative impact (dashed = indirect) Neutral F1.1: ➕/➖ F1.2: ➕ Fn: ➕ Fn: ➖ Fn: ➕/➖ • Proposition 1: Transparency is necessary condition for inter- organisational CI&D • F1.1 Important but not necessary • F1.2 Beneficial for synergy (efficiency, trust, mutual understanding 13 Results (RQ1)
- 14. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 14 What are risks and/or benefits of increasing transparency in an inter- organizational CI&D setting? Research Question 1 • Proposition 1: Increasing inter-organisational transparency of information is a necessary condition for inter-organisational CI&D • Proposition 2: Increased inter-organisational transparency of information is considered positive Results (RQ1)
- 15. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 15 Trust Inter-org. awareness Information sensitivity (privacy) Synergy (agile collab.) Inter-org. CI&D Legend Software dependent (direct) Software driven (indirect) Software agnostic Information transparency Positive impact (dashed = indirect) Negative impact (dashed = indirect) Neutral F1.1: ➕/➖ F1.2: ➕ F2.1: ➕ F2.3: ➖ F2.2: ➕ F2.2: ➕ F2.1: ➕ Shared project understanding F2.2: ➕ Fn: ➕ Fn: ➖ Fn: ➕/➖ • Proposition 2: Transparency of information is considered positive • F2.1: Positive for shared understanding and awareness • F2.2: Trust is increased • F2.3: Some information is sensitive Results (RQ1)
- 16. van der Valk et al. (via Eric Knauss) 17 Trust Inter-org. awareness Information sensitivity (privacy) Synergy (agile collab.) Inter-org. CI&D Responsibility split & IPR (inter-org.) harmonized tooling (Efficient) information sharing Legend Software dependent (direct) Software driven (indirect) Software agnostic Information transparency Positive impact (dashed = indirect) Negative impact (dashed = indirect) Neutral F1.1: ➕/➖ F1.2: ➕ F2.1: ➕ F2.3: ➖ F2.2: ➕ F2.2: ➕ F3.1: ➖ F3.1: ➕ F3.2: ➕ (perception of) physical distance F3.3: ➖ F3.2: ➕ F2.1: ➕ F6.1: ➕ Shared project understanding F2.2: ➕ Fn: ➕ Fn: ➖ Fn: ➕/➖ • Proposition 3: Typical risks of inter-organisational transparency (e.g. distance, tooling, IP) can be managed in practice. • F3.1: Reducing physical distance is beneficial • F3.2: Tooling can support information sharing • F3.3: Difficult to manage responsibilities and IPR Transparency and Contracts: CI/D in Automotive Ecosystem Results (RQ1)
- 17. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 19 Trust Inter-org. awareness Project quality Information sensitivity (privacy) Synergy (agile collab.) Inter-org. CI&D Responsibility split & IPR (inter-org.) harmonized tooling (Efficient) information sharing Legend Software dependent (direct) Software driven (indirect) Software agnostic Information transparency Positive impact (dashed = indirect) Negative impact (dashed = indirect) Neutral F1.1: ➕/➖ F1.2: ➕ F2.1: ➕ F2.3: ➖ F2.2: ➕ F2.2: ➕ F3.1: ➖ F3.1: ➕ F3.2: ➕ (perception of) physical distance F3.3: ➖ F4.1: ➕ F4.2: ➕ F3.2: ➕ F2.1: ➕ Shared project understanding F2.2: ➕ Fn: ➕ Fn: ➖ Fn: ➕/➖ • Proposition 4: A more open transparency policy improves the quality of the project and its results. • F4.1: Overall quality increased • F4.2: Short feedback loops beneficial Results (RQ1)
- 18. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 20 What are risks and/or benefits of increasing transparency in an inter- organizational CI&D setting? Research Question 1 • Proposition 1: Increasing inter-organisational transparency of information is a necessary condition for inter-organisational CI&D • Proposition 2: Increased inter-organisational transparency of information is considered positive • Proposition 3: Typical risks of inter-organisational transparency (e.g. distance, tooling, IP) can be managed in practice. • Proposition 4: A more open transparency policy improves the quality of the project and its results. Results (RQ1)
- 19. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 21 (How) can sufficient information be provided in inter-organizational CI&D collaborations? Research Question 2 • Proposition 5: Project members have sufficient information to perform their activities • Proposition 6: Information overload (in terms of frequent exchange) is unlikely to be considered a problem, if the exchanged information is precise. Results (RQ2)
- 20. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 22 Trust Inter-org. awareness Project quality Information sensitivity (privacy) Synergy (agile collab.) Inter-org. CI&D Responsibility split & IPR Information access (inter-org.) harmonized tooling (Efficient) information sharing Legend Software dependent (direct) Software driven (indirect) Software agnostic Information transparency Positive impact (dashed = indirect) Negative impact (dashed = indirect) Neutral F1.1: ➕/➖ F1.2: ➕ F2.1: ➕ F2.3: ➖ F2.2: ➕ F2.2: ➕ F3.1: ➖ F3.1: ➕ F3.2: ➕ (perception of) physical distance F3.3: ➖ F4.1: ➕ F4.2: ➕ F5.1: ➕ F5.2: ➕ F5.2: ➖ F3.2: ➕ F2.1: ➕ F6.1: ➕ Shared project understanding F2.2: ➕ Fn: ➕ Fn: ➖ Fn: ➕/➖ • Proposition 5: Project members have sufficient information • F5.1: Members have sufficient information • F5.2: Holistic project overview is needed • Proposition 6: Information overload is no problem • F6.1: Understandings through collaboration Results (RQ2)
- 21. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 23 Are contracts an impediment for scaling CI&D across company boundaries? Research Question 3 • Proposition 7: Strict contract-based collaboration can be seen as an impediment for inter-organisational CI&D Results (RQ3)
- 22. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 24 Trust Inter-org. awareness Project quality Information sensitivity (privacy) Synergy (agile collab.) Flexible contract Negotiation & competition Inter-org. CI&D Responsibility split & IPR Information access (inter-org.) harmonized tooling (Efficient) information sharing Legend Software dependent (direct) Software driven (indirect) Software agnostic Information transparency Positive impact (dashed = indirect) Negative impact (dashed = indirect) Neutral F1.1: ➕/➖ F1.2: ➕ F2.1: ➕ F2.3: ➖ F2.2: ➕ F2.2: ➕ F3.1: ➖ F3.1: ➕ F3.2: ➕ (perception of) physical distance F3.3: ➖ F4.1: ➕ F4.2: ➕ F5.1: ➕ F5.2: ➕ F5.2: ➖ F7.1: ➕ F7.1: ➕ F7.2: ➖ F3.2: ➕ F2.1: ➕ F6.1: ➕ Shared project understanding F2.2: ➕ Fn: ➕ Fn: ➖ Fn: ➕/➖ • Proposition 7: Strict contract- based collaboration: impediment for CI&D • F7.1: Flexible contracts beneficial for collaboration • F7.2: Traditional contracts better for negotiation Results (RQ3)
- 23. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 25 Are contracts an impediment for scaling CI&D across company boundaries? Research Question 4 • Proposition 8: Effects of adopting industry-wide standards and processes in a inter-organisational setting. Results (RQ4)
- 24. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 26 Trust Inter-org. awareness Project quality Information sensitivity (privacy) Synergy (agile collab.) Flexible contract Negotiation & competition Inter-org. CI&D Responsibility split & IPR Information access (inter-org.) harmonized tooling (adoption of) open source & standards (Efficient) information sharing Legend Software dependent (direct) Software driven (indirect) Software agnostic Information transparency Positive impact (dashed = indirect) Negative impact (dashed = indirect) Neutral F1.1: ➕/➖ F1.2: ➕ F2.1: ➕ F2.3: ➖ F2.2: ➕ F2.2: ➕ F3.1: ➖ F3.1: ➕ F3.2: ➕ (perception of) physical distance F3.3: ➖ F4.1: ➕ F4.2: ➕ F5.1: ➕ F5.2: ➕ F5.2: ➖ F7.1: ➕ F7.1: ➕ F7.2: ➖ F8.1: ➕ F8.2: ➕ F3.2: ➕ F2.1: ➕ F6.1: ➕ Shared project understanding F2.2: ➕ Fn: ➕ Fn: ➖ Fn: ➕/➖ • Proposition 8: Industry-wide standards and processes beneficial. • F8.1: Beneficial for information sharing. • F8.2: Maturity and management. Results (RQ4)
- 25. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 27 Trust Inter-org. awareness Project quality Information sensitivity (privacy) Synergy (agile collab.) Flexible contract Negotiation & competition Inter-org. CI&D Responsibility split & IPR Information access (inter-org.) harmonized tooling (adoption of) open source & standards (Efficient) information sharing Legend Software dependent (direct) Software driven (indirect) Software agnostic Information transparency Positive impact (dashed = indirect) Negative impact (dashed = indirect) Neutral F1.1: ➕/➖ F1.2: ➕ F2.1: ➕ F2.3: ➖ F2.2: ➕ F2.2: ➕ F3.1: ➖ F3.1: ➕ F3.2: ➕ (perception of) physical distance F3.3: ➖ F4.1: ➕ F4.2: ➕ F5.1: ➕ F5.2: ➕ F5.2: ➖ F7.1: ➕ F7.1: ➕ F7.2: ➖ F8.1: ➕ F8.2: ➕ F3.2: ➕ F2.1: ➕ F6.1: ➕ Shared project understanding F2.2: ➕ Fn: ➕ Fn: ➖ Fn: ➕/➖ Results (Overview)
- 26. van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 28 • Threats to Validity: Qualitative study with two companies • Key findings: More research needed! • Manage distance • Manage knowledge / information flow across organizations (transparency) • Suitable contracts needed • New collaboration models and empirical evidence about them are needed Discussion Transparency and Contracts: Continuous Integration and Delivery in the Automotive Ecosystem Eric Knauss Jacob Juul Rogardt Heldal Patricia Lago Patrizio Pelliccione Rob van der Valk ICSE 2018, May 31, Gothenburg, Sweden Contact speaker: eric.knauss@cse.gu.se | @oerich van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 27 Trust Inter-org. awareness Project quality Information sensitivity (privacy) Synergy (agile collab.) Flexible contract Negotiation & competition Inter-org. CI&D Responsibility split & IPR Information access (inter-org.) harmonized tooling (adoption of) open source & standards (Efficient) information sharing Legend Software dependent (direct) Software driven (indirect) Software agnostic Information transparency Positive impact (dashed = indirect) Negative impact (dashed = indirect) Neutral F1.1: / F1.2: F2.1: F2.3: F2.2: F2.2: F3.1: F3.1: F3.2: (perception of) physical distance F3.3: F4.1: F4.2: F5.1: F5.2: F5.2: F7.1: F7.1: F7.2: F8.1: F8.2: F3.2: F2.1: F6.1: Shared project understanding F2.2: Fn: Fn: Fn: / Results (Overview) van der Valk et al. (via Eric Knauss) Transparency and Contracts: CI/D in Automotive Ecosystem 30 Development Verifica1on Feature Func1on Electrical Architecture System ComponentParallel development: suppliers + In-house development “Big-bang” integra1on! Collabora1on based on contracts
Editor's Notes
- #2: Thank you for the introduction, Yvonne. It is my honor to present this work here at ICSE in my home town. And in fact, this was very intense work with continuous and transparent interaction with a great team of researchers. I know that Patrizio, Patricia and Rogardt are here at ICSE. Unfortunately, I also know that Rob is not here – a pity. Rob came to visit us in Gothenburg about two years ago, and I must admit that this required quite some persistence on his side to get all the help from us that he needed to setup the research - especially with a group of researchers being spread across Europé.. The project he investigated was in reality an experiment between Volvo and a trusted supplier to investigate new collaboration models. You can imagine it was also not easy to get interviews in such an exciting setup! So, it is safe to say, that without Rob, this work would not have been possible.