Harry Koehnemann Director of Technology 321Gang Applying the Scaled Agile Framework (SAFe) to Lean Systems Engineering

Transcription

1 Harry Koehnemann Director of Technology 321Gang Applying the Scaled Agile Framework (SAFe) to Lean Systems Engineering

2 Engineering Challenges Increasing complexity Rapid reduction in cycle times Risk meeting customer/market needs Products in a continuous release cycle Solutions cross organizational boundaries (Systems of Systems) System-wide collaboration demands (BOF vs. BOM) Increased product variation Compliance - contractual, regulatory 2

3 Our Current System Cannot Address Challenges System requirements, design, and schedule defined up front Difficult to defer decisions; reluctance to provide detail Project success defined by executing plan vs. delivering value Slow value deliver - delays, WIP, handoffs Not aligned to deliver value Big Design Up Front Development Test Value Knowledge 3

4 Where Does Agile Excel? Answer: Alignment and Collaboration But, only solves team-level alignment and collaboration Product Owner 24 Hours Product Backlog (Stories) Sprint Backlog (Stories) 30-day Sprint Potentially Shippable Software Sprint Planning Sprint Review (Demo) Sprint Retrospective 4

5 Apply Lean-Agile Principles #1-Take an economic view #2-Apply systems thinking #3-Assume variability; preserve options #4-Build incrementally with fast, integrated learning cycles #5-Base milestones on objective evaluation of working systems #6-Visualize and limit WIP, reduce batch sizes, and manage queue lengths #7-Apply cadence, synchronize with cross-domain planning #8-Unlock the intrinsic motivation of knowledge workers #9-Decentralize decision-making Leffingwell et al Scaled Agile, Inc. All Rights Reserved 5

6 Assume Variability; Preserve Options Aggressively evaluate alternatives. Converge specifications and solution sets. Allen Ward Flexible specifications Economic Trade-offs Design sets You cannot possibly know everything at the start Requirements must be flexible to economic design choices Designs must be flexible to changing requirements Preservation of options improves economic results Leffingwell et al Scaled Agile, Inc. All Rights Reserved 6

7 Integrate and Test Frequently Integration points control product development P A D C P A D C Integration points are pull events that accelerate learning Routine communication Reduce variation P A D C P A D C P A D C P A D C P A D C P A D C P A D C Objective evaluation Development can proceed no faster than the slowest learning loop The Lean Machine: How Harley Davidson Drove Top-Line Growth and Profitability with Revolutionary Lean Product Development Dantar P. Oosterwal Improvement comes through synchronization of design loops and faster learning cycles Leffingwell et al Scaled Agile, Inc. All Rights Reserved 7

8

9

10 Coordinate and Align Within and Across Layers Content Process/ Coordination Technical Integration /Test Solution Manager Value Stream Coordinator Solution Arch/Eng Systems DevOps * Product Manager RTE System Arch/Eng Systems DevOps * Product Owner Scrum Master Team 10

11 Organize Around Value Value doesn t recognize organizational or geographic boundaries Organize your people around your Value Streams Systems Value Streams deliver solutions Controls Mechanical 2015 Scaled Agile, Inc. All Rights Reserved 11

12 Programs Are NOT (necessarily) Value Streams Value Streams do not (necessarily) equate to contracts or programs Too much people motion leads to: Unpredictable team performance Lost productivity - form-storm-norm-perform Limited reuse Localized optimizations No economies of scale Don t bring people to the work Program 1 Bring work to the people Program 1 Program 2 Program 3 Program 2 Program 3 12

13 Use Value Streams to Deliver Product Variants Eliminate handoffs Organize around common product lines vs separate VSs Variants may be by feature or time (model year 17, 18) Variant A Variant B Variant C Variant A Variant C Variant B 13

14 Value Streams Deliver Solutions Feature Component Solution may be a component or feature Fund ARTs and cross-stream initiatives, not projects Vehicle Programs Powertrain HVAC Infotainment Active Safety Autonomous 14

15 Solution May Require a Context Represents Suppliers and Systems or Systems Customer continuously collaborates on multiple dimensions Content (backlog), technical, integration, program/budget Solution Manager Solution Arch/Eng Systems DevOps 15

16 ARTs Build Value Stream s Solution ARTs deliver each increment (and ideally sprint) Continually integrate solution at least each increment Value Stream roles coordinate content, technical, I&T Solution Manager Value Stream Coordinator Solution Arch/Eng Systems DevOps 16

17 Replace Gates with Cadence-Based Learning Cycles Base milestones on objective evaluation of working system Pull event to integrate entire system Work may include simulations, models, experiments, etc. Risk Risk Value Value Waterfall Lean-Agile 2015 Scaled Agile, Inc. All Rights Reserved 17

18 Learning Moves Variable to Fixed Simultaneously learn what we know, discover what we don t know Enablers create knowledge, decisions, and runway to build Features Decisions made with sufficient time to support feature building Accelerated by MBSE and Set-Based Design 18

19 Record Knowledge in System Intent Repository of collective system knowledge Single source of truth to communicate decisions Populated by results of Enabler work Facilitates impact analysis Supports regularity and contractual compliance 19

20 Use Models (MBSE) to Organize System Intent Customer System Constraints Regulatory Requirements Model (functionality, constraints) System Model (structure, behavior, simulation, parametrics, allocations) :Drive System Req Use Case Model :Vehicle :Accelerator :Radar :Steering :Brake :Vehicle Control :Motion Base Communication Impact analysis Strategic reuse Source for generating compliance documents :Chassis ctrl Domain Models (designs, analysis, etc.) MCAD Thermal, Software 20

21 Make Better Decisions with Set-Based Design Set-Based Design preserves options to make the best economic decisions based on objective evidence Multiple Design Options Set-Based Realizes the Cone of Uncertainty Learning Points Time Adjusted Single Option Too much to adjust No predictable stabilization Point-Based Leffingwell et al Scaled Agile, Inc. All Rights Reserved 21

22 Manage Change with Solution Kanban Organizes boards /CCBs Prioritized by Solution Management Reviewed by System Engineering (utilize System Intent) Analyzed by effected stakeholders Funnel Review Analysis Solution Backlog Solution Manager Change New Capability New Analysis Arch/Eng Teams ARTs SMEs 22

23 Connected Kanbans Backlogs contain centralized initiatives and local context Centralized strategy; decentralized decisions Increased visibility into the flow Hierarchical content governance system 23

24 Respect for people and culture Flow Innovation Relentless improvement Change Requires Leadership VALUE Lead the change Know the way; emphasize life-long learning Develop people LEADERSHIP People are already doing their best; the problems are with the system. Only management can change the system. Inspire and align with mission; minimize constraints Decentralize decision-making Unlock the intrinsic motivation of knowledge workers W. Edwards Deming Leffingwell et al Scaled Agile, Inc. All Rights Reserved 24

25 Acquire the Knowledge Implementing SAFe Leading SAFe 1 Ingrain deep SAFe knowledge (SPCs) Develop Lean-Agile leaders 2 Identify value streams; structure ARTs Train others SAFe Program Consultants (SPCs) Organize and support ARTs Implement Agile Portfolio Supporting role-based curriculum SAFe for Agile Teams & ART Quick start 4 Product Manager/ Product Owner Portfolio Management Scrum Master * Release Train Engineer * Organize and train Agile Teams Start the Train Plan and execute the first Program Increment SAFe ScrumXP Scrum Master & Product Owner Orientation 3 Leffingwell et al Scaled Leffingwell Agile, Inc. et All al. Rights 2015 Reserved Scaled Agile, Inc. All Rights Reserved 25

26 Harry Koehnemann Director of Technology 26

27 Rational Support For SAFe LSE CCM RM, DM, RELM QM Reporting Gang 27

28 DM/RELM System Model (structure, behavior, simulation, parametric, allocations) Links depend on needs (compliance, contract) RM NFR/Cust/Reg/Safety/etc. Requirements Use Case Model Domain Model (ECAD, MCAD, SW, etc.) System Req Simulation, Study, etc. Feature/Epic (lightweight business case) guides CCM Program Epic QM System Test Plan tracks PI Objective (Program) Feature/Enabler Program Test Plan Team Program/ Value Stream System Intent tracks PI Objective (Team) Story/Enabler Team Test Plan