Ontology-Based Knowledge Representation for Obsolescence Forecasting

Abstract

Sustainment refers to all activities necessary to keep an existing system operational, continue to manufacture and field versions of the system that satisfy the original requirements, or manufacture and field revised versions of the system that satisfy evolving requirements [3].The sales data is mainly in the form of number of units shipped. If it is not available, sales in market dollars or percentage market share may be used, as long as the total market does not increase appreciably over time [6].For some products, within the same type of the product, life cycle curves characterized by parameters k, μ, and σ can vary with some primary attributes of the product. Examples are memory chips whose life cycle curves vary with different memory sizes. Memory size is the primary attribute describing the memory chip that evolves over time [6][7][8]. For these products, if the primary attributes of the product are not considered, the parameters k, μ, and σ obtained from the sales data of the product are only average values for that product.The time range of the zone of obsolescence can be determined using data mining of historical data (e.g., lastorder or last-ship dates) to achieve more accurate obsolescence forecasting [8]. KeywordsCenter for e-Design, ontology, DMSMS, obsolescence, life cycle, forecast (2012) The impact and pervasiveness of diminishing manufacturing sources and material shortages (DMSMS) obsolescence are increasing due to rapidly advancing technologies which shorten the procurement lives of high-tech parts. For long field-life systems, this has led to an increasing disparity in the life cycle of parts as compared to the life cycle of the overall system. This disparity is challenging since obsolescence dates of parts are important to product life cycle planning. While proposed obsolescence forecasting methods have demonstrated some effectiveness, obsolescence management is a continuing challenge since current methods are very difficult to integrate with other tools and lack clear, complete, and consistent information representation. This paper presents an ontology framework to support the needs of knowledge representation for obsolescence forecasting. The formalized obsolescence forecasting method is suitable for products with a life cycle that can be represented with a Gaussian distribution. Classical product life cycle models can be represented using the logic of ontological constructs. The forecasted life cycle curve and zone of obsolescence are obtained by fitting sales data with the Gaussian distribution. Obsolescence is forecasted by executing semantic queries. The knowledge representation for obsolescence forecasting is realized using web ontology language (OWL) and semantic web rule language (SWRL) in the ontology editor Protégé-OWL. A flash memory example is included to demonstrate the obsolescence forecasting procedure. Discussion of future work is included with a focus on extending the ontology beyond the initial representation for obsolescence forecasting to a comprehensive knowledg...

Multi‐objective approach to forecast design refresh time due to COTS obsolescence

“…Skill and material obsolescence depend on the attributes of the system. There are several methods proposed to measure skill obsolescence 72–74 and material obsolescence; they depend on the risk factors and remedies with regard to the part and its technology 46,75–79 …”

Forecasting Obsolescence Risk and Product Life Cycle With Machine Learning

“…The most common short term reactive obsolescence resolution strategies include lifetime buy, last-time buy, aftermarket sources, identification of alternative or substitute parts, emulated parts, and salvaged parts [15][16]. However, these strategies are only temporary and can fail if the organization runs out of ways to procure the required parts.…”

“…Obsolescence risk forecasting generates a probability that a part or other element may fall victim to obsolescence [18][19][20][21]. Life-cycle forecasting estimates the time from creation to obsolescence of the part or element [12,16,22,23]. Using the creation date and life-cycle forecast, analysts can predict a date range for when a part or element will become obsolete [12,16,23,24].…”

“…Many methods have been developed to predict obsolescence of single feature products [12,16,23], for example flash drives. The flash drive may vary slightly in size and color but only has one key feature, memory.…”

Forecasting obsolescence risk and product lifecycle with machine learning

“…In practice, most firms do not have effective methods for predicting obsolescence and therefore are forced to over rely on reactive strategies [2]- [7]. The most common short term reactive obsolescence resolution strategies include lifetime buy, last-time buy, aftermarket sources, identification of alternative or substitute parts, emulated parts, and salvaged parts [15][16]. However, these strategies are only temporary and can fail if the organization runs out of ways to procure the required parts.…”