Product Design for Reliability
|Publication Date:||5 December 2020|
This business process describes designing reliability into specific products seeking PA that require engineering assessment from NR, or the Plant Assessment Body (PAB). It applies to existing and potential suppliers, and NR staff in departments including Engineering, Procurement, Supplier Quality Assurance (SQA), Reliability and Product Acceptance.
This business process applies to new and changed rail infrastructure products, systems, sub-systems, measurement equipment, materials, On- Track Machines and 'rail specific parts' of On -Track Plant entering the PA process. However, it is not mandated for On-Track Plant that was both designed prior to NR involvement and will be 'hired' , rather than purchased by NR, although it is recommended that this should follow the process where possible.
It also applies to products for use in new applications, with changes in product requirements and existing products sourced from new suppliers. It applies in a limited form when a supplier relocates their premises, or a product with an unchanged design is to be manufactured by a new supplier.
This business process is not mandated for PA requests exclusively covering new software or software changes, as demonstrating reliability of software typically requires different approaches to physical products. It does however, apply to any new or changed hardware related to, or containing such software.
NOTE: The lack of inclusion of software within this scope, does not reduce the need to continue to apply appropriate methods to the same extent as was expected prior to the introduction of DFR.
NR's requ irements should allow innovation by not being unnecessarily solution specific, but rigorous structured assessment and testing is still required.
To use part of an approved product, made by the same manufacturer, as a spare for itself, no DFR assessment will be required as part of the PA process.
In line with the Railways (Interoperability) Regulations 2011, any additional tests or requirements stated in this procedure for characteristics that are already defined in the TSI, do not apply and the assessor should request the certification instead, or ask the potential supplier to demonstrate they have met the essential requirements.
This business process does not outline the full set of PA requirements or entire design process. Other non-reliability specific steps are included for context. This process starts once the need for a product has been identified and ends once the product is being used in service with the confidence that it meets its required reliability performance, which either:
a) improves upon the reliability, availability or whole life cost of the item it replaces (or is no worse than a replaced obsolete product with the desired level of reliability);
b) has a net reliability benefit when implemented to address another issue; or
c) resolves an inherent safety issue even if the Mean Time Between Failures (MTBF) of any proven rapid interim fix would normally not be acceptable, followed by a redesign using this procedure that achieves both the safety and reliability requirements or a justification of why a redesign is not possible/desirable.
This business process integrates proven tools and methodologies into a potential supplier's existing design processes to create documented, traceable and controlled evidence of reliability, availability and maintainability. It helps enable compliance with BS EN 50126 before product acceptance (PA) is granted by NR and addresses train delay risk resulting from asset reliability.
Reliability tools and methods are now mandated and integrated into the design process in an effective sequence. This means that changes are identified earlier, add more value, are easier, more time efficient and more cost effective to implement.
This process focuses upon products, rather than geographical areas consisting of a range of products, although similar techniques can be applied.
The requirement definition steps (8.1.1 to 8.1.5) can help focus innovation in the correct d irection. Many of the subsequent steps are particularly powerful when applied to novel designs, to build confidence in the expected reliability when historic data may not exist.