The research question of this paper addresses how to achieve increased predictability in engineering. The research approach is constructive research. It is drawn on theoretical principles and ideas from Last Planner System and from production control in software engineering. Experience indicates that LPS is not adequate to achieve the desired predictability in engineering, but the underlying principles of involvement, continuous learning etc. are applied. The paper provides a solution based on a case study. The business of the case company is engineering, manufacturing and construction of mechanical installations for offshore oil and gas extraction and operations. The constructed solution is based on: • Delivery of drawings is part of the plan for manufacturing and construction. • How the delivery of drawings is met by the engineering department is controlled and planned in a separate process. • Dividing engineering work into phases. • Division of larger engineering objects into sub objects, meaning control areas. • The engineering control process focuses on control areas in the various phases. • A backlog of activities is created, from which tasks are prioritised into so- called sprints lasting for 1-4 weeks. A sprint may be the completion of a control area with a specific maturity in a phase. • The predictability in the sprint cycles is measured using PPC and causes of deviation.
Engineering, control, predictability, LPS, Scrum