Yanjun Qian

Overlapping and communication policies in product development

Shorter product life cycles in many industries impel firms to accelerate the product development process. Overlapping development stages, combined with frequent information exchange, is commonly regarded as a core technique for faster product development. However, overlapping and communication require additional resources and can be costly. We investigate the time-cost tradeoffs involved in concurrent product development to determine the optimal overlapping and communication strategies. The methodology was applied to a refrigerator development process in order to illustrate its utility.

Robust makespan minimisation in identical parallel machine scheduling problem with interval data

Parallel machine scheduling problems are commonly encountered in a wide variety of manufacturing environments and have been extensively studied. This paper addresses a makespan minimisation scheduling problem on identical parallel machines, in which the specific processing time of each job is uncertain, and its probability distribution is unknown because of limited information. In this case, the deterministic or stochastic scheduling model may be unsuitable. We propose a robust (min–max regret) scheduling model for identifying a robust schedule with minimal maximal deviation from the corresponding optimal schedule across all possible job-processing times (called scenarios). These scenarios are specified as closed intervals. To solve the robust scheduling problem, which is NP-hard, we first prove that a regret-maximising scenario for any schedule belongs to a finite set of extreme point scenarios. We then derive two exact algorithms to optimise this problem using a general iterative relaxation procedure. Moreover, a good initial solution (optimal schedule under a mid-point scenario) for the aforementioned algorithms is discussed. Several heuristics are developed to solve large-scale problems. Finally, computational experiments are conducted to evaluate the performance of the proposed methods.

A Novel Approach to DSM-Based Activity Sequencing Problem

Recently, there has been a growing interest in applying the design structure matrix (DSM) for planning projects that consist of many interrelated activities. One important objective of planning is to find an activity sequence so as to minimize the sum of superdiagonal numbers in a DSM. It is known that the problem is NP-complete and is difficult to solve. In this study, we prove several structural properties of the problem, and propose a heuristic for obtaining good feasible solutions. We find that based on the fold operation, a block of activities can be treated as a single activity. A novel hybrid algorithm is then presented for solving large activity sequencing problems. Finally, we perform a number of experiments and show that good solutions can be easily obtained by our approach. Moreover, the improvement achieved by the proposed approach is significant.

A fuzzy approach for sequencing interrelated activities in a DSM

Production and manufacturing systems often involve a myriad of interrelated activities. How these activities are organised and scheduled has a significant effect on the success of a system. Recently, the Design Structure Matrix (DSM) has been regarded as an effective tool for modelling and scheduling interrelated activities. Based on fuzzy set theory, this study explicitly addresses the uncertain activity dependencies in our formulation and develops a mathematical model for sequencing interrelated activities in a DSM. Because of the complexity of the model, a new approach, which embeds an exact algorithm within a framework of a local search heuristic, is presented for solving large problem instances. Testing results demonstrate that relatively good solutions can be easily obtained by our approach, thereby providing managers with an effective tool for scheduling a large number of interrelated activities with uncertain dependencies.

An effective approach for scheduling coupled activities in development projects

One of the greatest challenges in managing product development projects is identifying an appropriate sequence of many coupled activities. The current study presents an effective approach for determining the activity sequence with minimum total feedback time in a design structure matrix (DSM). First, a new formulation of the optimization problem is proposed, which allows us to obtain optimal solutions in a reasonable amount of time for problems up to 40 coupled activities. Second, two simple rules are proposed, which can be conveniently used by management to reduce the total feedback time. We also prove that if the sequence of activities in a subproblem is altered, then the change of total feedback time in the overall problem equals to the change in the subproblem. Because the optimization problem is NP-complete, we further develop a heuristic approach that is able to provide good solutions for large instances. To illustrate its application, we apply the presented approach to the design of balancing machines in an international firm. Finally, we perform a large number of random experiments to demonstrate that the presented approach outperforms existing state-of-art heuristics.

Organizing Interrelated Activities in Complex Product Development

An important decision problem faced by design managers is identifying an appropriate sequence of many interrelated activities. This study presents an effective tool for determining the activity sequence with minimum total feedback length in a design structure matrix. This problem can be modeled as the linear ordering problem with quadratic objective function and is extremely difficult to solve even for small problem instances. We first transform this problem into two equivalent linear integer programming problems, which can be solved exactly when the number of interrelated activities is not large. To provide good solutions for large problem instances, we further develop an exchange-based heuristic. Finally, the usefulness of the proposed methodology is validated through case study, computer experiments, and benchmark analysis.

Managing the Concurrent Execution of Dependent Product Development Stages

Concurrent product development, the practice of executing dependent product development stages simultaneously, has become the common mode of new product development because of the increasing importance of time-to-market. However, such simultaneous execution of dependent stages may substantially increase the total amount of rework. This research addresses the trade-offs involved in concurrent process, presents analytical models to determine the optimal priority ordering of initial development and rework, and the optimal overlapping duration. We first show that initial development is prior to rework when learning effect is taken into account. Then, based on the general assumption of nonnegative upstream evolution and the clear definition of maximum concurrency, we prove that the total development cost (including rework cost and opportunity cost of time) is convex with respect to the overlapping duration, and so the optimal overlapping duration can be identified by a simple binary search. After that we investigate the Pareto-optimal overlapping strategies for the cases where budget is given or the time to market is predetermined. Finally, the methodology is illustrated with a case study at a handset design company.

The trade-off between knowledge exploration and exploitation in technological innovation

Purpose This study aims to clarify the effect of team effort allocation between knowledge exploration and exploitation on the generation of extremely good or poor innovations. The influence of previous collaborative experience among team members on the effect of team effort allocation is also investigated to understand the relationship between team members’ collaboration networks and knowledge learning. Design/methodology/approach This study uses data of all patents granted by the US Patent and Trademark Office between 1984 and 2010. The inventors involved in a patent are regarded as members of the focal team. Logistic regression is used to analyze the data. Findings Allocating greater effort to exploration than to exploitation is beneficial to achieving breakthrough innovations despite the risk of generating particularly poor innovations. This benefit increases with collaborative experience among team members. Placing an equal emphasis on knowledge exploration and exploitation is not particularly effective in achieving breakthrough innovations; it is, however, the best strategy for avoiding particularly poor innovations. Originality/value This research not only provides valuable insights for research on innovation and knowledge management by studying the team effort allocation strategy used to achieve breakthroughs and avoid particularly poor innovations but also represents an advancement in bridging two streams of research – knowledge learning and social networks – by highlighting the influence of the team members’ collaborative networks on the effect of team effort allocation between knowledge exploration and exploitation.

Exact algorithms for the feedback length minimisation problem

Planning the sequence of interrelated activities of production and manufacturing systems has become a challenging issue due to the existence of cyclic information flows. This study develops efficient exact algorithms for finding an activity sequence with minimum total feedback length in a design structure matrix. First, we present two new properties of the problem. Second, based on the properties, we develop an efficient Parallel Branch-and-Prune algorithm (PBP). Finally, the proposed PBP is further improved by adopting hash functions representing activity sequences, which is referred as hash function-based PBP. Experimental results indicate that the proposed hash function-based PBP can find optimal solutions for problems up to 25 interrelated activities within 1 h, and outperforms existing methods.

Recent Advances in Concurrent Engineering Modeling

Over the last two decades, a number of studies have examined the trade-off involved in concurrent engineering (CE), time reduction versus additional effort for downstream rework. This study presents an overview of the recent CE modeling literature that examined this trade-off. We find that most CE models are built on the assumption that development stages are dependent where the principal information exchange between consecutive design stages is unidirectional, from upstream stage to downstream stage. According to literature review and field study, we believe such assumption is reasonable, because in many situations, current execution of design stages actually occurs within two sub-stages (Testing 1 and Development 2) which are sequentially dependent. In the future, we may also build analytical models based on interdependent stages so as to better understand the impact of project properties on best CE policies and product development performance.