Is AI testing beneficial for the manufacturer and social welfare? Optimal test strategy of a smart product

The rapid technological revolution of Industry 4.0 has set off rapid advancement in the smart product development. Testing the smart product's system before release to market is crucial for ensuring proper functioning. Smart product manufacturers must carefully consider different testing modes to choose the optimal one based on product characteristics and consumer's shopping intention. Additionally, different testing modes have varying impacts on the environment and social welfare, as well as profitability of the product. This study utilizes a two-period game theory model to analyze a smart product manufacturer's optimal testing mode and pricing decision. Three testing mode are evaluated: manual testing mode (MT mode), artificial intelligence testing mode (AT mode), and public testing mode (PT mode). Consumers' utility is influenced by factors such as consumer preferences, product quality, usage duration, and network externality effect. Our findings indicate that the optimal testing mode depends on quality improvement and network effect coefficients. Specifically, when the degree of quality improvement is high and the network effect is moderate, MT mode is preferred; when the quality improvement is high but the network effect is small, AT mode dominates; under other conditions, PT mode is optimal. Quality improvement and network effect coefficients have similar effects on pricing decisions but impact the profits in different manners. Social welfare is the highest under MT mode when the network effect is high and quality improvement is low, under AT mode when network effect is low and quality improvement is high, and under PT mode when both factors are high or low. Finally we provide a numerical example of how profit-driven business decisions may conflict with social welfare, potentially leading to lower social welfare than the theoretical optimum under the chosen testing mode strategy.