Engineering Process Management: Lecture 1 - Course Overview and Goals, Lecture notes of Management Theory

Download Engineering Process Management: Lecture 1 - Course Overview and Goals and more Lecture notes Management Theory in PDF only on Docsity! 1.264 Lecture 1 Course introduction Engineering process management Next class: Read Rapid Devt chapters 1-3. Hand in case study 1, 2 by noon before class Course outline SIL — George Kocur — Yin Wang No prerequisite; familiarity with Windows assumed Grading: — 10 homework sets (40%). Submit online, one per group. — Active learning (10%). Submit online, individually. ¢ Upload before (case studies) or after class (in-class software) — Midterm (25%) — Final exam (25%) Bring your laptop to class starting with lecture 6 — Work with a partner or by yourself Topics Engineering process change: rapid development methods — Develop, configure or manage software, business, and engineering projects — Unified modeling language (UML) for requirements, process modeling, communication among stakeholders Data modeling and databases — Business/system rules, normalization: database correctness — Query language (SQL) to build and query databases Web-oriented software process for development and configuration — Web technology: pages, documents, business rules, XML — Web services, service oriented architectures (SOA) Topics, p.2 * System architecture (software) — Servers, benchmarks, cloud computing — Mapping business needs to systems, databases ¢ Security process and software — Framework, protocols, attacks, secrecy/privacy, crime — Internet security ¢ Communications networks and services — Core technologies: fiber, wireless, CATV, satellite... — Networks: local, metro, wide area, enterprise networks — Protocols: Ethernet, TCP, IP, 3G/4G, ... Homework (project) ¢ Work in teams of two (1 and 3 allowed by exception) — Choose your partner this week. Ask TA to match you if you don’t find a partner. — You must get permission from the instructor to have a 1 or 3 person team ¢ Build a system for an aircraft parts distributor — First cycle of ‘spiral model’ of system development — Take 3 months to specify, design, prototype and assess ¢ And learn about the process and technologies — After first cycle, you could build an operational system ¢ Your prototype would almost be ok for a small operation ¢ If implementing a large supply chain (SCM) or transportation or other system, you follow the same steps A quick quiz What percentage of large projects have excess schedule pressure? — 25% 50% 75% a What percentage of small projects have excess schedule pressure? eb S07 re yi a C7 What percentage of large projects deliver on time and on budget? eb S07 re yi a C7 What percentage of large projects are cancelled or fail to deliver at I Wg e+ Eby 75% a LT) What staff increase is necessary to speed up a schedule by 25%? e+ Eby 75% a LT) How much are resource needs reduced by cutting project scope in lire Lit — 25% 50% 75% a How much have companies reduced time to market through better software practices in the last 10 years? LL) A 75% 100% ‘0 Answers to a quick quiz What percentage of large projects have excess schedule pressure? — 25% 50% 75% wT What percentage of small projects have excess schedule pressure? eb S07 rey a C7 What percentage of large projects deliver on time and on budget? eb S07 re yi a C7 What percentage of large projects are cancelled or fail to deliver at I Wg <b Eby 75% a LT) What staff increase is necessary to speed up a schedule by 25%? e+ Eby EW a LT) How much are resource needs reduced by cutting project scope in lire Lit — 25% 50% EWA a How much have companies reduced time to market through better software practices in the last 10 years? LL) WA 75% 100% i. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Results of the Standish Group CHAOS Report from 1994 to 2004. O Completed on time and within budget O Late, over budget, or with features missing O Canceled before completion Image by MIT OpenCourseWare. iy Steadily increasing use of spiral and other iterative models. Dimensions of development speed People — Matter the most: ability, motivation, management Process — Fundamentals, risk management, lifecycle planning — “Implement/code like hell” and chaos are still the most common approaches Product — Size and characteristics, phasing Technology — Product development/implementation environment -— Tools 15 People-Related Process-Related Product-Related Technology-Related Heroics Contractor failure Feature creep Silver-bullet syndrome Weak personnel Inadequate design Developer gold-plating Lack of automated source-code Wishful thinking Insufficient planning Requirements gold-plating Switching tools in the middle of a project Lack of user input Planning to catch up later Push-me, pull-me negotiation Overestimated savings from new tools or methods Undermined motivation Overly optimistic schedules Research-oriented development Noisy, crowded offices Code-like-hell programming Unrealistic expectations Insufficient risk management Lack of stakeholder buy-in Abandonment of planning under pressure Politics placed over substance Shortchanged upstream activities Adding people to a late project Insufficient management controls Lack of effective project sponsorship Wasted time during the fuzzy front end Friction between developers and customers Premature or overly frequent convergence Uncontrolled problem employees Omitting necessary tasks from estimates Shortchanged quality assurance Image by MIT OpenCourseWare. Technical fundamentals Spiral model as basis for development Cumulative cost Determine oes Identity and alternatives, and TeeaeRake constraints: Risk analysis Risk anelyais Evaluate Commit to an akarnatives approach for analysis. ‘Operational the next Risk on prototype iteration analysis..." Prototype 3 START we < Partition Requirements plan, lifecycle [Concept of ~~... madels, pian operation Softwaré-... benchmarks require- is sé ments Software Detaileit~- Perelopment Requirements product. 2esi9n ane validation design Code Plan the next Integration |Design validation : Unit iteration and test plan) and verification | best cIntegfation nd best pave “Acceptance evelop the test . ; f deliverables for the Release ! iteration and verify : that they are correct Image by MIT OpenCourseWare. 17