Intelligent Software Robotisation (ISR) of Work in Developing Economy Firms: A Complementarity and Configurational Analysis

Abstract

Intelligent software robotisation (ISR) is the use of software robots with intelligent capabilities, to reduce human intervention in business processes. The level of robotisation (LOR) is the degree of robot autonomy and intervention in a process and the two levels of general interest are the high LOR (high robot autonomy) and complementary LOR (shared autonomy with human oversight control). Currently typified by two technologies namely, robotic process automation (RPA) and artificial intelligence (AI), ISR brings a number of benefits to adopting organisations. These include cost savings, error reductions (accuracy), faster delivery times and increased productivity. In spite of the role that these ISR benefits (particularly cost savings and increased productivity), could play in helping firms in developing economies navigate the post-Covid economic challenges, regions such as, Africa have been noted for a slow ISR deployment rate. Furthermore, there are reports of a high failure rate of ISR projects in general. Motivated by these two issues, and adopting a configurational analytic approach, this study set out to understand the configuration of conditions that could explain the low ISR deployment rates and high project failure rates, with the ultimate purpose of developing a framework for successful ISR deployments in developing economies. A review of extant literature revealed some knowledge gaps related to the planned purpose, which helped further elaborate and firm up the study’s objectives. These were in the areas of ISR implementation approaches involving complementary LOR, employee job security concerns and coping strategies, as well as ISR theorization. Consequently, three objectives were developed to concurrently fulfill the purpose of the study while addressing these gaps. The three research objectives which steered this study are: (i) To examine the configurations of necessary and sufficient conditions which influence the decision to deploy intelligent software robotisation by a developing economy firm. (ii) To analyse the configurations of sufficient and necessary conditions which influence employees’ choice of coping strategies towards the deployment of intelligent software robotisation by a developing economy firm. (iii) To identify the nature of complementarity between humans and machines in intelligent software robotisation deployments in a developing economy firm. To address these objectives, the research efforts were guided by the critical realism paradigm, within which a parallel mixed methodology with a triangulation design was adopted. The quantitative inquiry employed a survey research method, while the case study method was used for the qualitative leg. The main consideration behind the choice of these methodological orientations, was their suitability for the research purpose which required a diverse and comprehensive set of data. Similarly, theoretical support was provided by three diverse models, on organisational technology adoption (TOE framework), IT coping behaviour and ISR complementarity. Data was collected from private sector organisations in the banking, insurance, telecommunications and business process outsourcing (BPO) sectors in Ghana. Data analysis was once again facilitated by diverse approaches namely, the Miles & Huberman (1994) approach and critical realist retroduction for the qualitative inquiry and Fuzzy-set Qualitative Comparative Analysis (FsQCA), a configurational analytic technique, for the quantitative inquiry. Findings from the first research objective indicate that, all the eight technological, organisational and environmental conditions investigated namely, Anthropomorphism, Technology Availability, Technology Accessibility, Firm Size, Organizational Structure, Industry Characteristics, Technology Infrastructure and Regulations, are necessary and sufficient conditions in influencing the decision to deploy ISR in developing economy firms. They could thus help to explain the low adoption rate and high failure rates of ISR initiatives in developing economies, which was among the primary motivations for this study. Conditions which emerged as requiring notable attention were the presence of technology availability, technology infrastructure (such as stable energy, broadband connectivity, data ecosystem, and cybersecurity) and industry characteristics (including labour union activities). These are valuable in explaining the low adoption rate of ISR in developing economies. Organisational structures (such as top management support, change management, skilled human resources and organisational culture) was also a notable finding, valuable in explaining the high rate of ISR project failures. The absence of regulations (such as laws on responsible & ethical AI and data privacy protection) as a core condition leading to a favorable decision to deploy ISR was a particularly noteworthy finding, in the light of recent global efforts towards regulations on the ethical and responsible deployment of AI. The finding is supported by literature which indicates that the presence of regulations could sometimes stifle the adoption of AI or other emerging technologies. This study’s qualitative inquiry cited the nascency of ISR adoption and industry-led self-regulation mechanisms as the reasons for the absence of regulations in the case country. The unique contributions from this objective are : the use of technology adoption research for a specific empirical analysis of the low adoption rates and high failure rates of ISR initiatives in Africa; the test of anthropomorphism in back-end software robots and labour union activities as constructs or conditions influencing ISR deployment, which had not been previously tested. Findings from the second research objective confirm the sufficiency and necessity of three conditions, Level of robotisation (LOR) of an ISR event (High or Complementary); Level of employee Control over working conditions, robots and emotions (High or Low) and the potential for self-interest fulfillment (Opportunity or Threat), in influencing employees’ selection of coping strategies towards ISR. The unique contribution from this objective was the test of LOR as a construct or condition in employees’ coping strategies which had previously not been tested. Findings from the third research objective, which were derived from the qualitative inquiry, were quite extensive (19 in all) and provided useful information on the various facets of the nature of complementarity in ISR deployments in developing economies. Some of the unique contributions from this objective included the cautionary approach adopted by firms towards the implementation of the level of intelligence in the robots and level of robotisation (LOR), fueled by concerns about internal skills maturity, system security and trust. Findings from the integration of the mixed methods inquiries (quantitative & qualitative) yielded a synthesis of information on a number of related themes from both inquiries: technology availability was integrated with awareness of ISR; organisational structure with top management support; employment security with labour union activities; regulations, anthropomorphism and LOR with themes of the same name in the counterpart studies. Finally, the originality of the study as a whole lies in the development of a proposed framework for successful deployment of ISR in developing economy firms, which brings together all the key findings from this study. The framework thus encapsulates this study’s contribution to research, practice and policy. For practice, the proposed framework provides guidance on general, human, machine and organisational conditions which influence the successful deployment of ISR in developing economies. On the flip side, the conditions are those which if ignored, could contribute to the failure of ISR initiatives. For research, the framework provides some new constructs and variables for the development of further models on ISR deployment. For policy, the general conditions segment of the framework, serves as a useful resource for areas of policy intervention.