Production Plan for Lucidcam

The production plan for Lucidcam was decomposed into two options: (A) NVIDIA chip and (B) Texas Instruments (TI) chip. As usual with product specification models, the key inputs are cost, quality, and risks. In the process of developing a comparative model for this case study, the decision for which manufacturer to support was heavily based on  maximizing the decision-maker Han Jin’s utility, where his utility was defined by likelihood of the company’s product reaching production on time. With the objective to have the product reach market by the set milestone date, the company should proceed with Design B, the Texas Instrument chip.  Although Design B had a higher risk of not meeting specification, in the end, it still had the higher likelihood of being able to manufacture on-time with a higher throughput than Design A.  Intuitively, it will make more sense to choose Design B, but we will utilize part of the decision analysis using utility analysis and decision tree as demonstrated below.

There were several events outside of our control, namely risks to the production timeline, that were dictated by the vendor’s production. These risks included the risk due to the complexity of design, risk to hardware failure, risk to firmware failure, and risk of supply chain failing. We quantified each of these risks and included them within our decision tree.

[pic 2]

Figure 1.1: Utility Profile

[pic 3]

Figure 1.2: Decision Tree

The optimal decision was based on the assumption that the profit margin and supply and demand were exempted from the analysis. Additionally, due to the limited resources of time and labor, it is inadvisable to move forward with a production plan utilizing both design plans simultaneously. Provided that an optimistic individual believes that there is a 50% probability of hardware risk for Design A will and that there is 10% probability of failure on firmware side.  Thus, a utility analysis will need to applied to decide on which design to adopt.  

To visually support this claim, a decision tree was designed to effectively mathematically represent the favorable outcome of Design B. LucidCam’s goal is to enter and stay in the market and be able to gain customer base and lead acquisition within the next six months, and the company will have higher likelihood if Design B was chosen.  Based on the objective of maximizing Han’s expected utility, we would select Design B which yields a higher expected utility given the uncertainties and utilities we defined in the model.

...