System engineering is a core discipline at HighPowerOptics. While we do not specialize in program management per se, the system engineering and program management roles have a lot of overlap and inter-dependence. Experience has taught us that good system engineering is enabled by good program management. Here we briefly summarize our philosophy of program management and system engineering.
HighPowerOptics, we manage projects with integrity and transparency. We approach every project with a sense of commitment, ownership and economy. We break down every challenge until we reach a level of simplicity that we can manage. We clearly communicate all assumptions and limitations. Some of the program management principles we live by are:
- Integrity is the fundamental basis for all engineering and interactions with the customer and/or end user. In nearly every project there is an element of discovery. Despite our best experience, insight and analysis, “we don’t know what we don’t know” – so we anticipate and plan for unexpected events.
- We break up the total project into separate phases that gradually advance the technical maturity of the product, with programmatic milestones and decision gates between phases. In general:
Phase I should develop the system architecture and validate the fundamental design concept with high fidelity analysis or software simulation (at least). Experimental validation should be used when high fidelity models are not available.
Phase II should retire risk to a mutually agreed level, generate a validated preliminary design, and result in a successful brassboard test or demonstration.
Phase III should refine the Phase II design, generate a critical design for a form, fit and function compliant prototype, result in a comprehensive test of the prototype, and produce a detailed production plan, including bill of materials and production cost model.
Phase IV should scale up manufacturing to low rate initial production. Feedback from prototype testing should be incorporated into the production design. Manufacturing processes should be completely documented. Production tooling, calibration and test equipment for initial low-rate production volume should be operational. Reliability testing on production units should be complete. Initial build lots must have passed quality assurance tests and customer acceptance tests with reasonable yield. The production cost model should be benchmarked with build lots of reasonable size.
- Whatever the budget size, we create a 10-20% management reserve at the beginning of each phase, judiciously scaling back the scope of work if necessary. We track actual costs and update spending forecasts on a weekly basis. “Nothing ever goes as planned.”
- We use multiple modes of communication when reporting back to the customer and/or end user: written (reports, tabular data), visual (graphs, photos, videos, drawings, and 3D renderings), and physical (sample products and models). Provide ample opportunity for question and answer.
- We report design deficiencies, test failures and/or programmatic (budget, resources, schedule, etc.) problems as soon as they are identified and confirmed. “Bad news does not age well.
Our technical area of expertise at HighPowerOptics is electro-optical systems. Dr. Winker has applied system engineering principles to a wide range of R&D projects, from low technical maturity, concept validation efforts to the development of space-flight qualified sensor hardware. System engineering methodology needs to be tailored to fit the complexity and budget of the project. Electro-optic subsystems can have considerable complexity, but they’re not nearly as complex as say, developing a new jet aircraft or a building a suspension bridge. Application of sound system engineering principles, combined with comprehensive testing throughout development, improves the end-product and reduces development cost. Here are some of the system engineering principles that we apply to the initial phase of electro-optics development.
- Understand the basic physics, engineering and operation of the end-product, and the environment in which it will be used.
- Determine all performance specifications, functional and interface requirements that could drive the design before starting the design. Capture all significant platform, environmental, schedule, and cost constraints as well. Identify the customer or end-user’s acceptance test criteria. Make sure that all temporary or undefined requirements have a “by when” date by which they will be finalized.
- Break the system down into manageable subsystems. Clearly understand the interplay between subsystems in terms of overall function, performance, reliability, safety, size and cost. Clearly define the modes of operation of each subsystem and their interfaces with each other and the outside world. Define all interfaces between subsystems in Interface Control Documents.
- Do appropriate system trade studies up front to identify the simplest system architecture and/or technical approach that meets the requirements. Apply the same trade study methodology to the development of subsystem and component designs.
- Develop a bill of materials early in the design, and feed recurring cost back into the trade studies to ensure that cost constraints are met. Identify where existing products can be used outright, or with minimal modification, to reduce subsystem development cost. Understand system performance trade-offs when off-the-shelf parts are used.
- Perform essential engineering analysis and modeling early that may be needed to derive component level requirements. Engage subject matter experts, if needed, to complete the requirements flow-down.
- Validate all new aspects of the design in the initial phase of the design process. Validate by experiment whenever high fidelity analytical or software models are not available. Ensure that test conditions are traceable back to operational requirements.
- Thoroughly document new processes and/or performance of new component technology and begin recording reproducibility data early.
- Assess risks thoroughly and manage they are retired early. Promptly update risk tables when risks are retired or become irrelevant, or new risks are identified.
- Assess manpower, facilities, cost and schedule of risk retirement and ensure adequate resources are available. Use Program Evaluation and Review Technique (PERT) 3-point analysis to estimate cost and labor.
- Verify compliance of the design with requirements. Verification by analysis or software modeling is acceptable when high fidelity models are available. Otherwise, experimental testing is the norm. “Test early and test often.”
- Allow enough time for reliability testing. Start early with engineering samples to identify weaknesses in the design. Find ways to test in the actual operating environment, or at least with combined effects of temperature, humidity and pressure.