Xiaozhou Meng at STC research center defended his doctorial thesis where he presents a new method that reduces maintenance costs of electronics in the automotive-, aerospace- and military industry. The research work has been to develop methods to cost-effectively manage obsolete components in embedded electronic systems.

When a product with expected long life cycle, such as an aircraft, is facing system maintenance and components with limited life cycle are no longer produced it creates problems for the industry.

- I have developed an optimization model for planning inventory purchases of components and to plan for the redesign of a system in which its components have different life cycles, says Xiaozhou Meng, PhD student at the STC Research Center, Mid Sweden University.

The model implies that the cost of maintenance of a system throughout its life cycle can be estimated for various technologies and platforms and in his thesis, Meng made ​​a risk analysis for embedded systems with long life cycles. Different platforms for hardware and software were evaluated based on different risk scenarios where components are assumed to exit the market.

An evaluation of this risk analysis shows that a system on FPGA platform with IP components independent of FPGA circuits manufacture and family, are the easiest to maintain, and thus provide the lowest cost for maintenance. This despite the fact that suppliers of FPGAs are keen to tie designers to their circuits brand by offering technology dependent IP modules at low cost.

- The study has been conducted in cooperation with Maxtimatec AB in Alfta and they have already in early stage seen that the model can reduce their costs significantly, says Meng.

The model consists of two stages: The first stage generates a graph of components that are expected to disappear from the market and measures to deal with the replacement of these. The graph is represented by a table used as input to a subsequent optimization in step two. In the second step, one can find the scheduling of steps leading to the lowest possible costs for maintenance of the system. For this optimization linear programming with integer models is used.

To reduce even more cost Mengs model should be used already when constructing products with long lifetime. To avoid unnecessary costs and create competitive advantages.

 

Xiaozhou Meng with his supervisors, opponent and examinations board.