A Time-Sensitive Knapsack Framework for Optimizing R&D Post-Competition Value
Open AccessResearch and development (R&D;) is a key source of value and competitive advantage. However, as decision-makers consider R&D; opportunities in practice, competition can often be a driving factor in determining opportunity value. Competition is a complex and interactive process that is difficult to predict and has significant effects on the value of R&D; investments over time. Initially promising investments may ultimately result in little value, due to the actions of others in a competitive environment. This is clearly reflected in the case of multiple organizations researching a new technology, in which one is first to patent and receives the majority of the rewards. These effects can be conceptualized as forms of time-sensitivity, in which the resulting values of R&D; opportunities are not static but instead are the dynamic product of interaction in an environment. Despite their importance, these effects are challenging to account for in the R&D; decision process. In this work, we establish a quantitative framework to determine effective decision-making behaviors for competitive conditions. The approach is built on the insight that decision-making can be optimized for the resulting Post-Competition Value (PCV) of opportunities, rather than for their initial pre-competition value. Our framework utilizes an agent-based model to simulate forms of interactive competition and determine PCV. Its fundamental mathematical structure is, in effect, a new time-sensitive formulation of the stochastic knapsack problem. Applying the framework, we evaluate candidate R&D; decision-making behaviors and characterize their performance in relevant competitive conditions. This defines the optimal strategy region for decision-making and correspondingly, the relative gains in PCV that can be attained in each competitive condition through improved decision-making. We synthesize the results into practical heuristics, which can be implemented by engineering managers to effectively account for competition in their R&D; decision process and gain a competitive advantage. Our heuristics are appropriate for typical R&D; applications. However, some real-world applications have complex or divergent competition dynamics that require specialized analysis. For these cases, we further introduce an evolutionary methodology that performs optimization of PCV within our framework in a model of a specific application. We demonstrate this methodology in three important defense R&D; applications, considering a case of defense industry companies, government laboratories, and nonprofits. In all cases, we identify optimized behaviors that achieve significantly more average value than standard alternatives that do not account for competition. We also show how the unique conditions in each case lead to major differences in the most effective decision-making behaviors. This analysis provides actionable recommendations to leaders in defense R&D;, and also a practical demonstration of how decision-makers can optimize for PCV in complex new applications of interest. Crucially, in both this methodology and our heuristic approach, implementation does not require a decision-maker to attempt to estimate post-competition value. Instead, our framework provides behaviors that already effectively account for competition effects through the underlying optimization.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.