Design for availability

Design for Availability is the design of a system considering availability or readiness as the major part of goal specification. The design parameters might include maintenance and inventory management, operation logistics, Prognostic Health Management (PHM) and reliability of system into the design process^[1]. The general purpose of this type of systems is in outcome-based contracts. Availability is a function of an item’s reliability (how often it fails) and maintainability (how efficiently it can be restored when it does fail)^[2]. Availability is a major factor of operational effectiveness along with performance of the system^[3]. Minimum required availability of complex system is a key factor of many distributed and repairable systems like ATM network or Airliner.

In Availability-based Contracts^[4] instead of parts, supplier is paid for a guaranteed level of services and performance a and system capability, like availability-based tarrif for electric power^[5]. The supplier often has to guarantee the availability and preparedness of system at lesser costs by considering the logistics as part of design. The contractor will also have more control over logistics and supply chain of system.

Recent interest in availability contracts that specify a required availability has created an interest in deriving system design and support parameters directly from an availability-based contracts. Because reliability and availability are both probabilities, there is a common assumption that the method used to compute system reliability based on component data can also be used for system availability calculations. A point to point allocation from parameter of operation and support to meet availability requirement in the performance space, given the required availability distribution is a direct way of design which must be followed by a sensitivity analysis and robustness of parameters. However, determining design parameters from an availability requirement is a stochastic reverse design problem. For safety, mission, and infrastructure critical systems, customers are often interested in buying the availability of a system through “availability contracts”, instead of actually buying the system itself. However, evaluating an availability requirement is a challenge for manufacturers and supporters of systems because determining how to deliver a specific availability is not trivial. The required availability can be defined over the instance of system or over the fleet. It can be defined over different time windows or in different geographical boxes. However availability optimization approaches provide solutions only at selected points in time (not all times), using mean time to failure (or fixed rate demand) and meantime to repair as deterministic values as part of convex optimization. ^[6]

While availability engineering in computer and network world is mostly about reducing unplanned downtime, reducing planned downtime is also important^[7]. Such planned downtime may include maintenance changes, operating system upgrades, backups, or any other activity that temporarily removes the application from service^[8]. this area of design is the subject of different discipline with different approach. in computer and network industry it is more targeting decreasing downtime where as in reliability world it might be seen as using higher reliability part in a cost effective manner.

Computer World Designing for availability is about anticipating, detecting, and automatically resolving hardware or software failures before they result in service errors, event faults, or data corruption — thereby minimizing downtime. The technology part of the solution requires multiple routes to application services and data. The operational part of the solution is to use only tested, proven processes (both automated and people-based) to support the application throughout its entire lifecycle.^[9]