Agile Systems Engineering & Agile at SKA Scale

Systems Engineering (SE) is the set of processes and documentation required for successfully realising large-scale engineering projects, but the classical approach is not a good fit for software-intensive projects, especially when the needs of the different stakeholders are not fully known from the beginning, and requirement priorities might change. The SKA is the ultimate software-enabled telescope, with enormous amounts of computing hardware and software required to perform its data reduction. We give an overview of the system and software engineering processes in the SKA1 development, and the tension between classical and agile SE.

Transcript

1. Agile Software Engineering & System Engineering at SKA scale Juande Santander-Vela SKA SW Systems Engineer SQUARE KILOMETRE ARRAY Exploring the Universe with the worlds’ largest radio telescope Astronomical Data Analysis
   Software & Systems (on behalf of the SE team)
2. SKA Key Science Drivers: The history of the Universe Cosmic Dawn (First Stars and Galaxies) Galaxy Evolution (Normal Galaxies z~2-3) Cosmology (Dark Energy, Large Scale Structure) Cosmic Magnetism (Origin, Evolution) Cradle of Life (Planets, Molecules, SETI) Testing General Relativity (Strong Regime, Gravitational Waves) Exploration of the Unknown Broadest range of science of any facility, worldwide
3. SKA Observatory Vision Exploring the Universe with the worlds’ largest radio telescope 3 sites 2 telescopes 1 observatory Design Phase: ~ €200M; 600 scientists+engineers, 80% complete SKA Phase 1 (SKA1) Construction: 2019 – 2025 Construction cost cap: €674.1M (2016 inflation-adjusted) Operations cost: (estimate) €89M/yr MeerKAT integrated Observatory Development Programme (€20M/year planned) SKA Regional centres out of scope of centrally-funded SKAO. SKA Phase 2: start mid-2020s ~2000 dishes across 3500km of Southern Africa Major expansion of SKA1-Low across Western Australia >50 years lifetime! Drives need for reliability, and adaptability
4. SKA Observatory Vision Exploring the Universe with the worlds’ largest radio telescope 3 sites 2 telescopes 1 observatory Design Phase: ~ €200M; 600 scientists+engineers, 80% complete SKA Phase 1 (SKA1) Construction: 2019 – 2025 Construction cost cap: €674.1M (2016 inflation-adjusted) Operations cost: (estimate) €89M/yr MeerKAT integrated Observatory Development Programme (€20M/year planned) SKA Regional centres out of scope of centrally-funded SKAO. SKA Phase 2: start mid-2020s ~2000 dishes across 3500km of Southern Africa Major expansion of SKA1-Low across Western Australia >50 years lifetime! Drives need for reliability, and adaptability
5. SKA Organisation Exploring the Universe with the worlds’ largest radio telescope ! Australia (DoI&S) " Canada (NRC-HIA) # China (MOST) $ India (DAE) % Italy (INAF) & Netherlands (NWO) ' New Zealand (MED) ( South Africa (DST) ) Sweden (Chalmers) * UK (BEIS/STFC) In discussion with: + Germany , France - Portugal . Spain / Switzerland 0 Japan 1 South Korea In the process of becoming an Inter- Governmental Organisation
6. SKA1 Design Consortia Exploring the Universe with the worlds’ largest radio telescope
7. SKA1 Design Consortia Exploring the Universe with the worlds’ largest radio telescope Software Heavy Elements
8. Have ~6-10 × FTE wrt SKAO SKA1 Design Consortia Exploring the Universe with the worlds’ largest radio telescope
9. 1000 km Map data ©2017 Terms of Use Filter Consortia Assembly, Integration & Verification Central Signal Processor Dish Infrastructure Australia Infrastructure South Africa Low-Frequency Aperture Array Mid-Frequency Aperture Array Science Data Processor Signal and Data Transport Telescope Manager Wideband Single Pixel Feeds Check/uncheck all GMT-7 GMT+12 Exploring the Universe with the worlds’ largest radio telescope 19 hour timezone span
10. SKA1-Low locations 1600 km N ➤ ➤ N Data SIO, NOAA, U.S. Navy, NGA, GEBCO Data SIO, NOAA, U.S. Navy, NGA, GEBCO Image Landsat / Copernicus Image Landsat / Copernicus SKA1-Mid locations 500 km N ➤ ➤ N Image Landsat / Copernicus Image Landsat / Copernicus Data SIO, NOAA, U.S. Navy, NGA, GEBCO Data SIO, NOAA, U.S. Navy, NGA, GEBCO SKA1 Sites Exploring the Universe with the worlds’ largest radio telescope SKA1-LOW: 50 – 350 MHz Phase 1: ~130,000 antennas across 65km SKA1-Mid: 350 MHz – 24 GHz Phase 1: 200 15-m dishes across 150 km
11. SKA1 Telescopes Exploring the Universe with the worlds’ largest radio telescope SKA1-LOW: 50 – 350 MHz Phase 1: ~130,000 antennas across 65km SKA1-Mid: 350 MHz – 24 GHz Phase 1: 200 15-m dishes across 150 km
12. SKA1 Notional Data Flow Exploring the Universe with the worlds’ largest radio telescope 2 Pb/s 7.2 Tb/s 7.2 Tb/s 8.8 Tb/s 8.8 Tb/s ~50 PFLOPS ~5 Tb/s ~250 PFLOPS Long Term Storage 300 PB/year LFAA Data Network Digital Data Backhaul Digital Data Backhaul CSP-SDP Network
13. SKA1 Notional Data Flow Exploring the Universe with the worlds’ largest radio telescope 2 Pb/s 7.2 Tb/s 7.2 Tb/s 8.8 Tb/s 8.8 Tb/s ~50 PFLOPS ~5 Tb/s ~250 PFLOPS Long Term Storage 300 PB/year LFAA Data Network Digital Data Backhaul Digital Data Backhaul CSP-SDP Network
14. SKA1 Notional Data Flow Exploring the Universe with the worlds’ largest radio telescope 2 Pb/s 7.2 Tb/s 7.2 Tb/s 8.8 Tb/s 8.8 Tb/s ~50 PFLOPS ~5 Tb/s ~250 PFLOPS Long Term Storage 300 PB/year LFAA Data Network Digital Data Backhaul Digital Data Backhaul CSP-SDP Network
15. SKA1 Notional Data Flow Exploring the Universe with the worlds’ largest radio telescope 2 Pb/s 7.2 Tb/s 7.2 Tb/s 8.8 Tb/s 8.8 Tb/s ~50 PFLOPS ~5 Tb/s ~250 PFLOPS Long Term Storage 300 PB/year LFAA Data Network Digital Data Backhaul Digital Data Backhaul CSP-SDP Network Similar to LOFAR in flexibility. See O4.2 by Adriaan Renting
16. How to develop so much software? Exploring the Universe with the worlds’ largest radio telescope
17. Agile Exploring the Universe with the worlds’ largest radio telescope
18. Agile, Scaled Agile Exploring the Universe with the worlds’ largest radio telescope (Framework)
19. SAFe for Lean Enterprises Large Solution SAFe ® Exploring the Universe with the worlds’ largest radio telescope Agile Teams, 5-7 people (including Scrum Master), ±2 people Agile Release Trains (ARTs) in Definition https://scaledagileframework.com/
20. Agile Manifesto • Individuals and interactions over processes and tools • Working software over comprehensive documentation • Customer collaboration over contract negotiation • Responding to change over following a plan Exploring the Universe with the worlds’ largest radio telescope http://agilemanifesto.org
21. The emphasis is on Quality Exploring the Universe with the worlds’ largest radio telescope Nick Rees talk I10.1 (Thursday 26th, 9:00 AM)
22. What about Systems Engineering? Exploring the Universe with the worlds’ largest radio telescope
23. INCOSE, 2004 “Systems engineering [SE] is an interdisciplinary approach and means to enable the realisation of successful systems […] [SE] integrates all the disciplines and specialty groups into a team effort forming a structured development process […] from concept to production to operation. [SE] considers both the business and the technical needs of all customers with the goal of providing a quality product that meets the user needs.” Exploring the Universe with the worlds’ largest radio telescope
24. INCOSE, 2004 “Systems engineering [SE] is an interdisciplinary approach and means to enable the realisation of successful systems […] [SE] integrates all the disciplines and specialty groups into a team effort forming a structured development process […] from concept to production to operation. [SE] considers both the business and the technical needs of all customers with the goal of providing a quality product that meets the user needs.” Exploring the Universe with the worlds’ largest radio telescope
25. INCOSE, 2004 “Systems engineering [SE] is an interdisciplinary approach and means to enable the realisation of successful systems […] [SE] integrates all the disciplines and specialty groups into a team effort forming a structured development process […] from concept to production to operation. [SE] considers both the business and the technical needs of all customers with the goal of providing a quality product that meets the user needs.” Exploring the Universe with the worlds’ largest radio telescope
26. INCOSE, 2004 “Systems engineering [SE] is an interdisciplinary approach and means to enable the realisation of successful systems […] [SE] integrates all the disciplines and specialty groups into a team effort forming a structured development process […] from concept to production to operation. [SE] considers both the business and the technical needs of all customers with the goal of providing a quality product that meets the user needs.” Exploring the Universe with the worlds’ largest radio telescope
27. INCOSE, 2004 “Systems engineering [SE] is an interdisciplinary approach and means to enable the realisation of successful systems […] [SE] integrates all the disciplines and specialty groups into a team effort forming a structured development process […] from concept to production to operation. [SE] considers both the business and the technical needs of all customers with the goal of providing a quality product that meets the user needs.” Exploring the Universe with the worlds’ largest radio telescope
28. David Long, CEO Vitech Corp “The role of Systems Engineers is to encode and translate engineering to all stakeholders.” Exploring the Universe with the worlds’ largest radio telescope
29. Nick Rees, SKAO Head of Computing “Systems Engineers capture and remind us of all the bullshit stakeholders say about the system.” Exploring the Universe with the worlds’ largest radio telescope
30. Exploring the Universe with the worlds’ largest radio telescope
31. Exploring the Universe with the worlds’ largest radio telescope
32. Exploring the Universe with the worlds’ largest radio telescope
33. Juande Santander-Vela, SKAO Software SE “Systems Engineers have to provide useful engineering artefacts to stakeholders, notice still missing definition, and supply it ASAP” Exploring the Universe with the worlds’ largest radio telescope Freshly coined for #ADASS2017CL!
34. Juande Santander-Vela, SKAO Software SE “Systems Engineers have to provide useful engineering artefacts to stakeholders, notice still missing definition, and supply it ASAP” Exploring the Universe with the worlds’ largest radio telescope Or assess and record the risk of not doing it Freshly coined for #ADASS2017CL!
35. INCOSE, 2004 “Systems engineering [SE] is an interdisciplinary approach and means to enable the realisation of successful systems […] [SE] integrates all the disciplines and specialty groups into a team effort forming a structured development process […] from concept to production to operation. [SE] considers both the business and the technical needs of all customers with the goal of providing a quality product that meets the user needs.” Exploring the Universe with the worlds’ largest radio telescope With an assessment of what are the risks against quality and user needs
36. SE must be Agile, or it provides diminished value Exploring the Universe with the worlds’ largest radio telescope
37. Agile Manifesto • Individuals and interactions over processes and tools • Working software over comprehensive documentation • Customer collaboration over contract negotiation • Responding to change over following a plan Exploring the Universe with the worlds’ largest radio telescope http://agilemanifesto.org
38. Agile SE • Keep communication with all stakeholders involved in the project, with speedy and clear dissemination of agreements • Continuous integration of the engineering artefacts… and code, preferably self-documented • Get devs (and SEs) that are invested, not just contracted → Gain‑share frameworks • Have systems & processes that accommodate system (and code) change and prove requirements are satisfied → don’t do document- centric requirements management Exploring the Universe with the worlds’ largest radio telescope
39. Agile SE • Keep communication with all stakeholders involved in the project, with speedy and clear dissemination of agreements • Continuous integration of the engineering artefacts… and code, preferably self-documented • Get devs (and SEs) that are invested, not just contracted → Gain‑share frameworks • Have systems & processes that accommodate system (and code) change and prove requirements are satisfied → don’t do document- centric requirements management Exploring the Universe with the worlds’ largest radio telescope At all phases! Including design, construction, commissioning, and operations… Don’t disband SE!
40. Agile SE • Keep communication with all stakeholders involved in the project, with speedy and clear dissemination of agreements • Continuous integration of the engineering artefacts… and code, preferably self-documented • Get devs (and SEs) that are invested, not just contracted → Gain‑share frameworks • Have systems & processes that accommodate system (and code) change and prove requirements are satisfied → don’t do document- centric requirements management Exploring the Universe with the worlds’ largest radio telescope At all phases! Including design, construction, commissioning, and operations… Don’t disband SE!
41. Agile SE • Keep communication with all stakeholders involved in the project, with speedy and clear dissemination of agreements • Continuous integration of the engineering artefacts… and code, preferably self-documented • Get devs (and SEs) that are invested, not just contracted → Gain‑share frameworks • Have systems & processes that accommodate system (and code) change and prove requirements are satisfied → don’t do document- centric requirements management Exploring the Universe with the worlds’ largest radio telescope At all phases! Including design, construction, commissioning, and operations… Don’t disband SE!
42. Agile SE • Keep communication with all stakeholders involved in the project, with speedy and clear dissemination of agreements • Continuous integration of the engineering artefacts… and code, preferably self-documented • Get devs (and SEs) that are invested, not just contracted → Gain‑share frameworks • Have systems & processes that accommodate system (and code) change and prove requirements are satisfied → don’t do document- centric requirements management Exploring the Universe with the worlds’ largest radio telescope Partners, not contractors/suppliers mentioned by JC Guzman on I2.1 At all phases! Including design, construction, commissioning, and operations… Don’t disband SE!
43. Agile SE • Keep communication with all stakeholders involved in the project, with speedy and clear dissemination of agreements • Continuous integration of the engineering artefacts… and code, preferably self-documented • Get devs (and SEs) that are invested, not just contracted → Gain‑share frameworks • Have systems & processes that accommodate system (and code) change and prove requirements are satisfied → don’t do document- centric requirements management Exploring the Universe with the worlds’ largest radio telescope Partners, not contractors/suppliers mentioned by JC Guzman on I2.1 At all phases! Including design, construction, commissioning, and operations… Don’t disband SE!
44. Agile-ish Practices at SKAO • SE planning • We have a full plan, from which the backlog is constructed • Progress reviewed every 2 weeks, with forward look to another 2 weeks • Systems Modelling • Integrated System Model for SKA1_Common and SKA1_Low in Cameo SysML • Second model for SKA1_Mid in Vitech CORE • Used for ICDs, discussions • SKA1_Mid and SKA1_Low Block Diagrams • Used for reference, updated in agile way, but low priority Exploring the Universe with the worlds’ largest radio telescope
45. Agile-ish Practices at SKA1 • Consortia work-planning • Agile for SDP, TM • LFAA uses agile for software development of AAVS1 prototype • Others want to embrace SAFe for construction, just not now • Multimodal communication • Email, SMS, Confluence, JIRA, Jama comments, Slack, Vidyo, Skype… Exploring the Universe with the worlds’ largest radio telescope
46. Not so Agile-ish Practices at SKAO • Engineering Change Process (ECPs) and Change Notes (CNs) • Bug, or Feature? Still TBD for ECPs → more time to reflect • Need for formality/signature/buy-in across consortia requires time; we need to be able to move from as agreed to as designed before as signed… but how? • Lack of JIRA/backlog culture → starting to improve • Lack of self-awareness of performance metrics → starting to improve Exploring the Universe with the worlds’ largest radio telescope
47. Agile SE is not for free… • Coordination/communication takes time! • Self-organisation frees time from management, but needs assistance from monitoring → back to coordination! • Automation of monitoring and communication can help… but still the effectiveness of the communication needs to be monitored. Exploring the Universe with the worlds’ largest radio telescope a lot of But it makes it easier to focus on the right things
48. Conclusions • Agile SE can be seen as a tailoring of SE; you still need to identify which are the processes that provide value to your system/organisation. • It is a challenge to create teams that can work together, but making it on a cadence makes it easier to achieve it… through repetition. • We have to improve a lot at SKAO… but we’re working on it. • Agile for Agile’s sake is a fad; you become Agile it to build quality, and respond to mistakes, sooner. Exploring the Universe with the worlds’ largest radio telescope
49. Thanks! Exploring the Universe with the worlds’ largest radio telescope SQUARE KILOMETRE ARRAY
50. Questions? Exploring the Universe with the worlds’ largest radio telescope SQUARE KILOMETRE ARRAY j.santander-vela@skatelescope.org