Working With Robots: Human Resource Development Considerations in Human–Robot Interaction

Attitudes Toward Robots

How people perceive robots is critical for HRI because negative attitudes can make employees avoid robots (Nomura et al., 2006). Although public opinions about robots in the workplace are mostly negative due to the threat to jobs and roles (Gnambs & Appel, 2019), employees’ feelings about their robot partners tend to be more favorable than the public’s perceptions, in general (Kahn et al., 2011). Attitudes toward robots are likely to depend on each employee’s individual background, experiences, and abilities (Hancock et al., 2011). For example, technology savvy and robot-literate individuals are more open to having interaction with robots (Bartneck et al., 2007; Soh et al., 2020).

Research has also paid considerable attention to trust, which focuses on how well people accept robots and believe that they can be fully committed to a robotic system to achieve a goal (Kim et al., 2020; Soh et al., 2020). In the workplace, the importance of trust lies in transparent collaboration in which employees willingly recognize the information that robots provide, accept suggestions from robots, and predict interactions (Hancock et al., 2011). Although robot performance is the main factor affecting trust in HRI (Kim et al., 2020), there are also human-related and environmental factors such as individual experiences and expertise, group tasks for collaboration, and organizational policies related to integrity and safety (Natarajan & Gombolay, 2020; Sanders et al., 2019).

Another group of researchers has focused on anthropomorphism in which transmission of human-like attributes to robots may increase the acceptance of robots (Damiano & Dumouchel, 2018) and improve collaborative performance (Fraune, 2020; Natarajan & Gombolay, 2020). While positive attitudes are expected to reduce employees’ cognitive workload and lead to more commitment to HRI, overreliance can undermine the value of HRI (Natarajan & Gombolay, 2020; Rahman & Wang, 2018). The role of HRD should be to evaluate the effects of employees’ attitudes about HRI and ensure a positive experience when working with robots through organizational activities and interventions such as orientation, small pilot tasks, virtual training, and job opportunities.

Technology Readiness

Given that a new industrial method is typically applied based on users’ capabilities, employee readiness is a key consideration in HRI (Charalambous et al., 2017; Gombolay et al., 2017). Parasuraman (2000) defined technology readiness as preparedness and willingness to accept and use new technologies to achieve goals at work. With low readiness, employees are likely to feel insecure about and uncomfortable with new technologies, which could lead to low motivation and productivity (Islam et al., 2019; Parasuraman & Colby, 2015). In contrast, high readiness could help employees have a positive view of technology and efficiently manage it for their jobs. Individual characteristics and organizational support are key elements in enhancing technology readiness (Blut & Wang, 2020).

Charalambous et al. (2017) developed a systematic model for technology readiness in HRI based on NASA’s evaluation scale (Mankins, 2009). In their model, low levels of readiness are related to identifying and recognizing the basic principles and concepts of robots. Middle levels of readiness include a clear understanding of the process and complexities of HRI, allocating the necessary resources, and understanding the place of employees in the actual workplace of HRI. A high level of readiness is where actual production begins, and employees are given discretion over operations and handling system errors. Given that failed technology implementation is often caused by a lack of human capabilities (Abdelaal et al., 2020; Hentout et al., 2019), it is crucial for organizations to provide appropriate training, organizational support, and senior management involvement to increase employees’ level of readiness.

Communication With Robots

Effective HRI requires a clear understanding of the behaviors and anticipation of the intended purposes for both humans and robots (Hentout et al., 2019). Thus, humans and robots need mutual communication. Robotic research has focused mainly on robots’ capability to be aware of and react to the intent of the human operators through sensory and voice-command devices (Makrini et al., 2017). The development of AI and machine learning (i.e., a method to improve AI through experience and data) enables robots to learn better. Through recent advancements, robots are starting to instantly and accurately recognize their human partners’ intent and appropriately respond to variations in human behaviors (Oliff et al., 2020). However, fluent mutual communication between humans and robots enhances the effectiveness of HRI in a similar way to open communication among humans, which is a key for team performance (Villani et al., 2018).

For effective communication, employees are expected to play two roles. One is to be aware of and learn robots’ behaviors (e.g., actions, work processes, and outcomes) and the other is to teach and guide robots by providing demonstrations, feedback, and reinforcement (Sheridan, 2016). In their experimental study, Nikolaidis et al. (2015) revealed that mutual learning about the roles humans and robots should play increased the performance of a collaborative task. Bidirectional human–robot communication led to improved mutual adaptation for humans and robots in their collaborative tasks compared to single-direction communication (Nikolaidis et al., 2017). Therefore, developing effective communication skills with robots, which is different from a human-to-human setting, will be a new but key area of HRD for HRI.

Human–Robot Team Building

Human–robot teams are expected to recognize the team capacities of different units (humans and robots) and establish ways to improve group processes based on key elements of team building including goals, roles, and member relations in the team. Based on team building components, first, goal setting for human–robot teams may require identification of the individuals’ capabilities, task allocation, and resources for collaborative tasks (Gombolay et al., 2017; Sheridan, 2016). In addition, humans and robots have different skills and abilities that should be considered in individual and group goal setting (Bruno & Antonelli, 2018). Although workload balance may not be a critical issue in human–robot teams, too heavy or too light of a human workload can have a negative impact on individual well-being and team performance (Gombolay et al., 2017). The decision making for collaborative tasks should be based on the definitions of the tasks and human–robot team capacity (Tsarouchi et al., 2017).

For role clarification in a human–robot team, individuals should understand their own and other team members’ (humans and robots) tasks and duties (Hentout et al., 2019). Team communication should be a two-way interaction so both humans and robots are clearly aware of the other in terms of roles, work progress, and upcoming objectives in collaborative tasks, including learning robots’ jobs, teaching human jobs to robots, and role switching between humans and robots (Nikolaidis et al., 2017; Whitsell & Artemiadis, 2017). Although role-clarification activities may not lead to a positive experience for humans (e.g., cognitive and emotional pressure), they may help build trust between humans and robots through improved communication (Nikolaidis et al., 2015). Human–robot team performance also increases when members proactively assist, guide, and work together with other members rather than reactively participating in collaboration only when needed (Baraglia et al., 2017).

Human relationships with robots relate to teamwork skills and synergy. Fraune (2020) found that although humans tend to favor humans over robots, humans accept and treat robots as in-group members when they work together. Like in human teams where relationships among members depend on how they interact with each other, research has indicated that active human–robot communication and strong human trust are crucial factors that influence human relationships with robots and eventually increase the performance of human–robot collaboration (Hentout et al., 2019; Nikolaidis et al., 2015). Organizational support and interventions to adapt to working with robots will help employees develop relationships with robots and learn how to treat their robot partners.

Leading Multiple Robots

To effectively lead a team with multiple robots, it is important to solve problems that occur within the team just like in a human work team. In a human–robot mixed team, problems between humans and robots may be addressed by ensuring proper execution of goal setting, role clarification, and relation development. The performance of a team that involves multiple robots depends on the leader’s role, which is likely more salient when the robots are different from each other (Prewett et al., 2010). Several researchers have also emphasized the importance of human skills in multirobot systems because of the complex variations that cannot be anticipated or pre-treated (Adams, 2009; Hirche & Musić, 2017; Oliff et al., 2020). To effectively deal with problems that impact performance, a team with multiple robots requires agile approaches, such as establishing the decision-making process, delegating problem-solving actions, and developing robot management skills (Abdelaal et al., 2020; Moniz, 2014). HRD approaches to leading multiple robots may include enhancing employees’ leadership and management skills for robots through training programs and learning sources and providing relevant experiences.

Systemwide Collaboration

HRI involves far more complicated processes and activities than operating machines or using fully automated robots. For effective management and development of HRI, organizations are encouraged to practice multifaceted approaches by reconsidering the whole work system including decision processes, flexibility, guidelines, and policies (Moniz, 2014). In this regard, van Wynsberghe and Li (2019) proposed a novel approach for operators (employees), developers (engineers), and management to help them intensively and systematically collaborate, optimize effectiveness, and safeguard employees. In HRI, an essential consideration is likely to be the transparent incorporation of various stakeholders’ characteristics (Vrontis et al.), such as the operator’s experiences and needs, the developer’s expertise and design intention, and the management’s system perspective and production objectives. For effective systemwide collaboration, HRD should play a coordinator role that links the various actors and aspects. Coordinating functions may include observing robots and their performance, sharing feedback with robot developers on robot operation, developing employee competency for more effective collaboration with robots, and ensuring the well-being and satisfaction of employees (Thomas et al., 2018).

Safety Interventions

Safety is another important consideration in a workplace with robots. Unlike a conventional setting with machines and large industrial robots that mainly require physically isolating employees from unexpected contact with the machines, HRI requires more specific and complex management because physical contact is necessary (Benos et al., 2020). To take full advantage of employees’ performance and to maximize the effectiveness of HRI, safe interactions should be guaranteed (Villani et al., 2018). The field of HRI has recognized the importance of protecting employees against accidents and injuries and, therefore, has devoted considerable attention to designing safe and reliable operation environments and systems. The efforts related to safety guidelines include International Organization for Standardization (ISO) standards that are specific to industrial environments (ISO 10218: 2011), personal use (ISO 13482: 2014), and collaborative robots (ISO/TS 15066: 2016). Literature has also sought to design systems (e.g., control systems and motion/collision detection) to ensure collaboration without sacrificing operation speed and separation from robots (Saenz et al., 2020).

The causes of accidents in HRI can be categorized as engineering errors, human errors, and poor environmental conditions (Vasic & Billard, 2013). Although the last two categories are due to human factors (e.g., inattention, fatigue, neglecting safe procedures, inadequate training programs, and incorrect procedures), the literature has focused predominantly on the first category: engineering errors. Few studies have dealt with a human-centered perspective of safety, and the focus has been limited to employees’ perceptions of robots (Zacharaki et al., 2020). Given the need for organizational effectiveness based on safety interventions (Robson et al., 2012), HRD should contribute to and take the lead in developing and conducting safety interventions, especially when risk exposure severity is high (Burke et al., 2011). For example, safety training sessions and safety awareness tools help employees learn and practice safe behaviors. In addition, a psychological safety assessment will allow organizations to monitor employees’ safety concerns and feelings about working with robots. Well-established systems, standards, and employee awareness will also help avoid safety-related disasters, enhance employees’ trust in working with robots, and result in better collaborative performance (Maurtua et al., 2017; You & Robert, 2018).

Ethical Issues

Robot ethics has received increasing attention in various disciplines including engineering, computer science, law, psychology, and philosophy. The aspects of robot ethics have mainly focused on how to design, deploy, and treat robots, which requires the development of moral competences (Malle, 2016). Most studies on robot-related ethics have focused on the human impact of social and domestic robots (Lin et al., 2011; Smids et al., 2019). However, the ethical implications of HRI in the workplace have not been sufficiently explored. Thus, there is little understanding of what human and organizational issues could occur when there are immoral practices or ignorance of appropriate actions in collaborating with robots (Fletcherd & Webb, 2017).

I identified two approaches to address ethical concerns in HRI. First, organizations need to establish ethical systems and environments to protect employees in HRI. Employees should have sufficient information about robots and full awareness of the code of behavior and risks (Fletcherd & Webb, 2017). Employees who have little trust in robots should not be pushed into HRI work settings because distrust may overwhelm them and deteriorate their well-being (Kim et al., 2020; Rahman & Wang, 2018). In addition, the data that robots generate and collect should be stored and used ethically with a rigorous protocol to protect privacy (Lee et al., 2011).

The second approach is that organizations should guide employees to properly treat and manage robots. A clear standard for working with robotics helps employees understand appropriate behaviors toward robots and their own human rights and responsibilities in HRI (Lin et al., 2011). Organizations should also be aware of immoral applications of robots, such as inappropriate teaching/programming for malevolent purposes, hacking other robots, sharing unnecessary sensitive information with robots, and social loafing when employees shift their work to robots (Bonaci et al., 2015). HRD should lead initiatives for HRI ethics by educating and training employees and be resourceful in establishing an ethical culture and system in the organization.

Psychological Capital

Working with robots introduces attitude challenges in human–robot relations. For effective communication and collaboration with robots, employees need positive attitudes toward robots. Organizations with employees who are reluctant or unprepared to interact with robots may need organizational accommodations and developmental remedies. Given that a positive mental state is related to in-group approaches and relationships (Avey et al., 2011; Chen et al., 2019), human perceptions and behaviors related to HRI correspond to psychological capital theory. Psychological capital requires psychological resources (e.g., confidence, optimism, and resilience) and promotes growth and development of individuals, all of which contribute to employee well-being and performance (Luthans, Youssef, & Avolio, 2006). The role of HRD in promoting psychological capital and positive employee behaviors involves building a culture of trust, promoting reciprocity, and implementing ethical standards in the organization (Luthans et al., 2006). Although psychological capital originally focused on employees’ work settings, HRI requires a new application of the process. Specifically, for employees with traditional tools, HRD has focused on improving the culture and interactions among employees. However, as humans increasingly work with robots as colleagues in HRI, it has become more important to focus on the significance of human perceptions toward robots. The same importance of developing a positive culture in human–human relations also applies to human–robot relations. Through trusting, ethical, and safe interactions with robots, employees will be confident, optimistic, and resilient as they work with robots, robotic systems, and human–robot relations, which, in turn, will result in effective HRI. Thus, an expanded theory that considers human–robot relations can be newly added in HRD which will help scholars and practitioners better understand how positive attitudes are shaped in HRI and how HRI affects the workplace.

Workplace Learning

Collaboration with advanced AI-based robots suggests a mutual learning model that includes how employees communicate, learn, and train their collaborative robots. To date, HR scholarly interest in AI technology has focused on how to utilize machine learning for HR activities, such as assessing individuals, predicting performance, identifying training needs, measuring training effectiveness, and identifying teamwork patterns (Garg et al.; Mulder & Beer, 2020). This approach exemplifies how scholars view technology as a supporting tool in Bennett’s (2014) typology. However, the reality of HRI is the assumption that there are mutual influences between humans and robots that involve more than humans using machines. Recent scholarly attention has focused on the trainer’s (human employee’s) role in an HRI environment since robot intelligence improves as robots learn from humans. In human-to-robot training, Ramos et al. (2020) underscored that the training capabilities and subject-matter expertise of humans are critical for machine learning and building AI models. Learning about robots includes not only obtaining skills and knowledge about robotic technology, but also understanding their behaviors for appropriate reactions (Sheridan, 2016). The mutual learning model may open up a new learning realm that relates to developing human competencies for learning and training robots and an AI improvement process of teaching, coaching, and guiding robots.

Organizational Learning

The broad impact of HRI on HRD suggests that scholars and practitioners need to rethink learning at the organizational level. The core of organizational learning is adjusting to a change in the organization at the individual, group, and organizational levels and fostering knowledge creation and learning (Argote & Miron-Spektor, 2011; Nonaka, 1994). Learning experiences from a new technological system such as HRI will restructure the existing knowledge and help create a new culture and structure (Popova-Nowak & Cseh, 2015). However, unlike a traditional organizational learning model in which information is obtained and evolves from internal activities among individuals, organizations with HRI may require different approaches to learning. These organizations may seek to extend their technological knowledge and learning capabilities from external sources, such as M&A, outsourcing, and talent acquisition (Argote & Miron-Spektor, 2011). Given that sources of state-of-the-art knowledge like robotic technology are very scarce, organizations may rely on a broad range of online expert communities or publicly shared resources on the internet (e.g., YouTube) to increase robot technology readiness and obtain more up-to-date knowledge despite the security and validity risks (Bean, 2019).

Virtual HRD (i.e., a digital environment to improve the expertise and performance of employees and organizations) requires non-traditional sources of learning through informal methods and plays an important role in organizational learning about technology (Bennett, 2014). Similarly, with increased HRI, strategic, smart, and legitimate uses of technology-mediated learning have become critical for businesses (Thite, 2022). Therefore, HRD research should pay increased attention and establish new theories for organizational learning practices related to new technology and related challenges. For HRI, organizational learning processes are more complicated because HRI knowledge is created by both employees and robots through their interactions, and the knowledge is interpreted and analyzed by multiple stakeholders (e.g., employees, engineers, programmers, managers, HRD practitioners) to transfer the knowledge to other work units and projects. Organizational learning in HRI should not only be flexible but also robust to create and safeguard information obtained from numerous sources.

Organization Systems Model

From an organization development perspective, AI and HRI robotic technologies, as a push factor of change, impact major design components of organizational systems (Nadler & Tushman, 1980). Technologies are likely to affect the individual- or group-level performance, and the impact of the fundamental changes brought on by HRI affects the whole organization. This paper identified key HRD considerations for HRI at the individual, group, and organizational levels. At the individual level, a new competency model of employees is required given the knowledge, skills, and abilities needed for HRI. Ensuring job enrichment of these employees is a critical concern because HRI may affect employee satisfaction and well-being. At the group level, goal setting and task design approaches may change because of the new work boundaries between humans and robots and because the definition of group membership is becoming blurred. At the organizational level, the need for systemwide collaboration among operators, developers, and management requires a novel organizational structure. Human resource systems require new mechanisms for deploying, developing, and assessing both humans and robots. The inter-relatedness among the HRD considerations within levels and across levels indicates that HRI affects the whole organizational system. Thus, the organizational changes and impacts of HRI should be understood through the lens of the systems model.