Abstract:

With the rapid development of information technology and the leading wave of digital transformation， many enterprises begin to transform operations such as online-onsite collaborative operation mode， based on the digital technologies of cloud computing， artificial intelligence， and digital platforms. In the context of digitalization， online-onsite collaborative operation mode will be widely applied with the integration of platforms and tools， and the creation of intelligent gadgets， in order to realize lowering costs and boosting productivity. For example， some tasks in projects including installation， inspection， testing， etc.， are more suitable for online-onsite collaborative mode， that is， technological experts execute guidance online by visualization through information equipment.Here， an extension of multi-skill resource-constrained project scheduling problem （MSRCPSP） is proposed and investigated in this study. The distinct feature of collaboration between online employees and onsite employees is considered in MSRCPSP. More specifically， the human resources are categorized into online workforce who can only perform knowledge-based skills and onsite workforce with both knowledge-based skills and operation-based skills. The project activity network is consisted of two types of activities： 1） purely online activities which can only be executed by online workforce； 2） onsite activities which can be done in two ways， one is executed by onsite workforce alone， the other is executed by collaboration of online and onsite workforce. Compare with the traditional MSRCPSP， there are new decision variables in this problem for reasonably selecting execution mode for each onsite activities， the purely onsite mode， or the online-onsite collaboration mode. For the purely onsite mode， the onsite workforce who meet the conditions is scarce so there exists a delay on the spatial transfer and it is hard to implement parallel multitasking operations. For the online-onsite collaboration mode， although the resource waiting time is alleviated， the operational efficiency is highly relied on the collaboration between online workforce and onsite workforce especially for the challenging activities， which will cause the extension of processing time. Consequently， with the goal of minimizing the project completing time， a programming model is proposed to gain the optimal activity execution modes， scheduling scheme and workforce assignment under the constraints of resource limitation， skill type and level requirements， precedence relations. An adapted genetic algorithm embedded nine strategies is designed based on "highest level of skill first"， "the most average workload first" and "the shortest processing time first"， to satisfy the computational requirements of large-scale problems.The instances for different problem scales， namely 10， 20， 30， 60， 90， 120， are randomly generated. The large number of experiments are conducted， the model is validated and the proposed genetic algorithm has proven to be effective compared with Gurobi. Besides， the results are analyzed and several managerial insights are drawn. First， an increase in the proportion of online activities contributes to the shortening of the processing time， while an attention should also be paid to the overloaded online activities. Second， whether to adopt a collaborative mode should be fully deliberated including the durational amplifying effect of the collaboration difficulty， the scarcity of online and onsite resources， and the disparity in the knowledge-based level between online and onsite resources. Third， the frequencies of transfer are influenced by the ratio of onsite resources， so onsite resources should be added appropriately to reduce the flow frequency between regions when the transfer time is lengthy.This research can effectively offer guidance on the assignment of online and onsite workforce for projects with the rapid development and application of various digital technologies in many enterprises. Besides， it extends the research paradigm and theoretical underpinnings of the MSRCPSP.

Key words: project scheduling, personnel assignment, online-onsite collaboration, multi-skilled employees, genetic algorithm