Creation of collision data base through the “ bottom-up ” approach

The present epoch of knowledge economy, research and innovation poses unpredictable challenges for high-tech enterprises in aerospace industry. Knowledge and technology have mostly become the key source of creating competitive advantage. However production of successful aerospace products and service is impossible without collisions. The aim of the paper is to elaborate a bottom-up approach that allows reducing collisions by setting up a database of collisions (CDB) that occurs in the life cycle process. The introduced definition of “collision” determines any breach in the implementation process, leading to a deviation in product quality. The classification of collision types is given and the stages of collision management system in aerospace enterprises are described. By means of Risk Priority Numbers (RPN) method the authors have proposed to evaluate the significance of probable production collisions. According to conclusions of the analysis, collision database allows to accumulate, formalize and store invaluable experience of skilled practitioners, especially in unique production process. Other research outcome deals with the understanding that collision database allows detecting business processes where collision occurrence is not always obvious.


Introduction
The production of rocket and space equipment assumes a large variety of industrial processes and manufacturing operations, which are characterized by their specificity: a high level of workability, a single production, high risks (financial, economic, environmental, etc.), high level of responsibility.Reliability of aerospace technology ensures due to quality engineering and manufacturing processes.
Moreover the Russian aerospace enterprises are continuously updated technological support: there are additional options, progressive design decisions, etc.In aerospace engineering and manufacturing the enforcement requirement is considered to be the continuity and knowledge transfer.The problem of continuity is achieved through the creation of a knowledge management system (KMS) in the aerospace company.
The creation of the KMS from scratch is quite costly because these costs will be recouped over a long period of time.However, space services are now very much in demand.The intensity of the space launches is increasing every year.Table 1 presents data of Russian space launches from 2010 to 2016 [1].The presented data suggest that the production of the satellite can be regarded as mass production.The presence of failed launches demonstrates the need for recording, collecting and processing errors, so called collisions.Reaction velocity to a collision occurred must be sufficiently high so that all admitted ones can be resolved immediately, i.e. in the manufacture of the next product.

Main part
The paper offers a pragmatic and cost saving approach to the practical realization of the collision data base creation.The proposed approach can be used as small and medium enterprises involved in aerospace projects and large-scale enterprises as well, through existing corporate information systems.
One more accepted perspective in KM area is knowledge influence to innovation growth.This perspective is extremely important in the high R&D intensity sectors which require radically new innovation approaches.
Knowledge acquisition can influence the R&D performance and it's capability to convert resources into innovative products best explains differences in enterprises' innovation performance [2].
Although the above mentioned perspectives are thoroughly elaborated there are many vague areas in KM.One of them is the warehousing and registering collisions in production process based on bottom-up approach.
Even typical and well organized production process could face unpredictable risks and obstacles.Hence the growing interest in high-tech enterprises in recent years associated with the creation of knowledge management system that would have solved the global problem of information formalization.But the process of implementing KMS for high-tech enterprises is rather complicated due to a number of factors: • Corporate information system (CIS) that brings together developers, designers and production staff on a single platform are rather expensive.
• Creating internal portals that facilitate internal information exchange do not guarantee the quality of products.
• Particular characteristics in processes of unique production creates certain difficulties in the allocation and description of typical algorithms for design, engineering and production.
• There is an acute problem of the experience transfer (young workers lack the intuitive and practical experience, lack of mentoring, etc.), so that once there is a problem of formalization of experience and intuition.
Professional experience and intuition usually appear the job age.Thus one could pose a question what the professional experience is?Our opinion is that this is the knowledge of previous errors, failures, collisions and consequences to which they lead.In these terms skilled practitioners are more competitive than the novices (inexperienced employees).So any high-tech enterprise faces a problem how to transfer knowledge from experienced professionals to younger employees?In order to find out a solution of experience and intuition transfer, a concept of collision has been introduced.
Collision is considered as any breach of production process, leading to a deviation in product quality (performance/ non-compliance of required output parameters, etc.).
We discover a collision when aerospace system or its elements: -doesn't (don't) function at all; -doesn't (don't) function in a definite period of time; -functions (function) in not a proper way; -functions (function) not in full.The cost of making changes to the project, on average, is hundreds of times less than the cost of making changes to the finished product.Early detection and prevention of collisions at their occurrence place lead to significant cost savings for the enterprise as a whole.
Creation of collision data base (CDB) designed for such conditions will allow formalizing intuitive decisions, both the correct and incorrect, and their consequences, and also warehousing professional experience of employees and preventing failures in production process (fig.1).

Figure 1. Collision database creation
In other words, if the aerospace enterprise hasn't opportunity to create a global knowledge management system that integrates information, experience and intuition (a "top-down" approach), it could accumulate and formalize experience in collisions and their solutions (a "bottom-up" approach).
The "bottom-up" approach allows creating and disseminating knowledge at the operational level, and only after the successful implementation of innovative proposals one can generalize the results to the enterprise level.The proposed approach also takes into account those accumulated production of knowledge held by more experienced employees (fig.2).

Figure 2. Main characteristics of two approaches
One of the key advantages of the "bottom-up" approach is the timely detection and prevention of production errors that occur due to the lack of sufficient professional experience.Throughout the product lifecycle, from design to the finished product, there is a lot of different changes that have a different nature arise, and may adversely affect the quality of the product.
The first step in creating CDB will systematize all possible collision types and their place of occurrence (Table 2).The third type of RPN method is used for various process control levels of an aerospace enterprise, administration and production: the first level integrates administrative system, information system, human resources management, the last level includes the organization of the basic working units.

Analysis
While using the Risk Priority Number (RPN) method to assess risk of consequences, the analysis team must: • Determine the type of collision; • Rate the severity (S), the likelihood of occurrence (O), and the likelihood of prior detection (D) for each collision • Calculate the RPN by multiplying three ratings: RPN = Severity x Occurrence x Detection The parameters S, O, D range from 1 to 10.Thus, the maximum value of 1,000 RPN.The classification is carried out for the serious consequences of each type of collision shown in table.5.
The RPN method based on expert evaluations shows that human factor collisions may have more negative impact than technological collisions.
The next step in CDB management identifies all the potential collisions in order to assess the more critical ones.As discussed, a collision can be considered as: (i) the real lack of capacity to perform a certain action, or (ii) the presumed/perceived inability to perform it.

Table 2 .
Classification of collision typesThe example of a process analysis is given in the table 4.

Table 5 .
Short Example of RPN method application in collision management