Assembly/Disassembly Optimization Model (The "Salvage" Model)

[Description] | [Inputs] | [Outputs] | [Features] | [Known Bugs] | [Walkthrough] | [References]

Description

A methodology that incorporates simultaneous consideration of economic and environmental merit during the virtual prototyping phase of electronic product design has been developed.  The model allows optimization of a product lifecycle that includes primary assembly, disassembly, and secondary assembly using a mix of new and salvaged components. Optimizing this particular lifecycle scenario is important for products that are leased to customers or subject to product takeback laws. Monte Carlo simulation is used to account for uncertainty in the data, and demonstrates that high-level design and process decisions may be made with a few basic metrics and without highly specific data sets for every material and component used in a product. A web-based software tool called "Salvage" has been developed that implements this methodology.

The Salvage tool allows the exploration of the primary assembly, disassembly and secondary reassembly of new systems using selected parts from the disassembled primary unit. The basis for the exploration is cost, quality (yield), waste material, and consumed material.  The model computes the characteristics of the primary assembly and testing operation (cost, yield, waste, material consumption), the cost of system disassembly, and the characteristics of assembly and testing of a secondary assembly that uses a combination of new and salvaged parts.

Inputs

Three types of inputs are supported by this application:

1) Component/Material Acquisition – descriptions of all the components used to create the primary and secondary assemblies of the product under design.  Examples include: chips, connectors, heat sinks, and printed wiring boards. The following characteristics can be input:

• Primary component cost, quality (yield), waste material generated when the component was originally fabrication, and material consumed when the component was originally fabricated (per component)
• Cost of disassembly (per component)
• Quality after buy back  (per component)
• Damage during disassembly (per component)
• Test cost and test effectiveness (per component)
• Whether or not the component is salvaged for the secondary assembly

2) Manufacturing/Assembly – the process of assembling the system (the same process is assumed for both the primary and secondary assemblies).  The following characteristics can be input:

• Description of the component mix for each process step
• Attach cost, defects introduced, material wasted, material consumed (Assembly Steps)
• Test cost and test effectiveness (Test Steps)

3) Disassembly/Salvage – special characteristics of the disassembly and take back process.  The disassembly process can be custom created by the user or defaulted to the reverse of the assembly process. The following characteristics can be input:

• Fraction of primary parts that are returned and are salvageable
• Ratio of secondary build quantity to primary build quantity
• Buy back cost

Outputs

The following outputs are available for the product under design:

• Primary assembly cost, quality (yield), waste material, material consumed
• Disassembly cost
• Secondary assembly cost, quality (yield), waste material, material consumed
• Bar charts showing the probability distributions associated with any of the above outputs

Features

Special features associated with this application include:

• Option to characterize all input values as distributions (uniform, normal, triangular, lognormal)
• Monte Carlo analysis control
• Editors for process and components
• Full set of help pages

Known Software Bugs

The following bugs are known to exist in the present version.

1. Use the right mouse button when clicking on the tabs to change forms (the left mouse button is unreliable).
2. Always add or remove rows from tables before you make changes to the data in the cells. Changing forms will lock in all data changes.
3. The units on the mean and standard deviations printed on the bar charts are MKS ($, fraction, kg).
4. The "Confidence level" input on the Setup dialog box is not used at this time.
5. The scatter plots are not completely operational at this time.
6. The tables columns may become jumbled when scrolled horizontally.  Changing forms will properly refresh the table.

Walkthrough

A walkthrough presentation containing screen dumps from this application can be viewed: View walkthrough

References

P. A. Sandborn and C. F. Murphy, "A Model for Optimizing the Assembly and Disassembly of Electronic Systems," IEEE Trans. on Electronics Packaging Manufacturing, vol. 22, no. 2, April 1999, pp. 105-117 .

C. F. Murphy, C. Mizuki, and P. A. Sandborn, "Implementation of DFE in the Electronics Industry Using Simple Metrics for Cost, Quality, and Environmental Merit," in Proc. IEEE International Symposium on Electronics & the Environment, May 1998.