Ubiquitous tools, connected things and intelligent agents: Disentangling the terminology and revealing underlying theoretical dimensions
First Monday

Ubiquitous tools, connected things and intelligent agents: Disentangling the terminology and revealing underlying theoretical dimensions by Katrin Etzrodt and Sven Engesser



Abstract
Research on the social implications of technological developments is highly relevant. However, a broader comprehension of current innovations and their underlying theoretical frameworks is limited by their rapid evolution, as well as a plethora of different terms and definitions. The terminology used to describe current innovations varies significantly among disciplines, such as social sciences and computer sciences. This article contributes to systematic and cross-disciplinary research on current technological applications in everyday life by identifying the most relevant concepts (i.e., Ubiquitous Computing, Internet of Things, Smart Objects and Environments, Ambient Environments and Artificial Intelligence) and relating them to each other. Key questions, core aspects, similarities and differences are identified. Theoretically disentangling terminology results in four distinct analytical dimensions (connectivity, invisibility, awareness, and agency) that facilitate and address social implications. This article provides a basis for a deeper understanding, precise operationalisations, and an increased anticipation of impending developments.

Contents

Introduction
Review of terminology
Ubiquitous Computing
Internet of Things
Smart Objects and Environments
Ambient Intelligence
Artificial Intelligence
Disentangling the terminology
Extraction of underlying dimensions
Conclusion

 


 

Introduction

The past few decades have witnessed profound changes in computer technology, which fundamentally affect life and society. These changes are sometimes compared to the industrialisation of the nineteenth century (Greengard, 2015; Siemoneit, 2003). Santucci [1] put the development in a nutshell: “[...] in the nineteenth century, machines learned to do; in the twentieth century, they learned to think; [...] in the twenty-first century, they are learning to perceive they actually sense and respond”.

Not only machines but also everyday objects (e.g., pens, cups, plates, and mirrors) are increasingly equipped with sensors, processors, and actuators. They are able to communicate with both humans and machines alike, and they actively affect their environment. Consequently, objects develop an agency of their own and have the potential to become social actors, which implies that handling these objects becomes social acting. Despite this considerable extension of social life, the social implications of the current technological changes have been widely neglected. A reciprocal exchange between technological and socio-scientific disciplines rarely occurs.

One reason for this dearth of research may be the lack of scientific terminology and theoretical concepts for the technological changes and their social implications. Science has not yet sufficiently developed an appropriate language for this highly relevant object of research. The extant terminology has been largely restricted to arcane circles of specialists or specific disciplines. It has also been frequently used in an ambiguous und unsystematic manner.

This paper aims to provide a common vocabulary and theoretical framework for the current changes in computer technology and their social implications. We have gathered the most relevant terms from the literature, disentangled them semantically, uncovered their underlying theoretical dimensions, integrated them in a conceptual model, addressed their social implications, and accessed them for a variety of disciplines. Consequently, this paper prepares the ground for the systematic socio-scientific analysis of the current technological changes, while enabling the transfer of knowledge beyond disciplinary boundaries.

 

++++++++++

Review of terminology

Drawing on extant review articles (Dhawan, et al., 2016; Olson, et al., 2015; Friedewald, et al., 2010), we have identified scientific terminology referring to current changes in computer technology. We focused on terms that describe applications in everyday life (e.g., conversational assistants or vacuum cleaner robots), excluding infrastructures and technological ontology (e.g., the Semantic Web or System of Systems). In addition, the terms were required to be relevant to the discussion and research of the field. Therefore, we examined terms for their occurrence in journal articles, conference proceedings, and other types of publications with the help of the Scopus literature database. As a threshold, terms needed to appear in a minimum of 1,000 articles. Thus, we have limited the terminology to the five most common conceptual frameworks: Ubiquitous Computing, Internet of Things, Smart Objects or Environments, Ambient Intelligence, and Artificial Intelligence.

We analyzed the substance of these concepts in three steps. First, we identified the 10 most cited texts for each term in Scopus, giving a total of 50 publications. However, only seven provided a distinguished depiction of the underlying concepts. Second, we studied additional literature about the concepts taken from review articles, the seven most cited articles, and from discussions on the subject with computer scientists. This approach made it possible to include important monographs and other publications that we did not discover the first time. Finally, the chronological evolvement and connections with other terms were explored using the Scopus analysis tools and their keyword collection. However, due to the lack of space, we have limited the scope of this article to the elaboration and presentation of the knowledge thus gained. The following section will present the key questions of the conceptual frameworks, their core aspects, and their relation to each other.

 

++++++++++

Ubiquitous Computing

Key question. How can computer technology blend in the natural human mode of dealing with tools?

Core aspects. The term Ubiquitous Computing [2] and its conceptual framework originate from Mark Weiser and his work at Xerox PARC in the 1990s (Want, 2009). This builds upon the fundamental critique of the personal desktop computer claiming that its restriction to one location, its cryptic operability, and its focus on assistance-features draw too much attention to the technology instead of to the task itself (Weiser, 1994; Friedewald, et al., 2010). In contrast, Ubiquitous Computing aims at enhanced “computer use by making many computers available throughout the physical environment, while making them effectively invisible to the user” [3]. The two most important features of computer technology in this regard are invisibility and omnipresence.

Invisibility. Drawing on the work of Heidegger and Simon, Weiser (1991) argued that tools should disappear from consciousness to enable their users to concentrate on the task. Technology should migrate into the background (Friedewald, et al., 2010). Inspired by Suchman (1985), as well as Lave and Wenger (1991), Weiser (1993) suggested that this could be achieved through adaptation to different work contexts and through simplicity of use. He conceptualised this kind of invisibility on physical and operational levels.

Physical invisibility refers to easy-to-use computer tools resembling non-technological objects in terms of appearance, use, and purpose. Through these tools, human-computer interaction can become intuitive [4]. The first projects in Xerox PARC focused on three types of devices (Want, 2009): Tabs (small handheld devices) and Pads (book-sized devices) were designed for individual and mobile use. They were the prototypes of smartphones and tablet computers. Liveboards (big stationary screens), in contrast, were designed for teamwork in face-to-face situations, which were the prototypes of digital whiteboards.

Operational invisibility refers to computer operations that “weave themselves into the fabric of everyday life until they are indistinguishable from it” [5]. Consequently, Ubiquitous Computing focuses on adapting to different contexts of use, such as locations, situations, tasks or user behavior (Mattern, 2003). It aims at autonomously switching between the periphery and center of attention, depending on the context of the task (Weiser and Brown, 1997).

Omnipresence. This conceptual framework inaugurated the modern era of computer technology, in which many computers are used by one person (Want, 2009). Computer devices are supposed to populate the workplace like ordinary tools, such as “notes, titles on book spines, labels on controls, thermostats and clocks, as well as small pieces of paper” [6]. The concept envisions the co-existence of “hundreds of computers in a room” [7]. They should be physically and operationally invisible and they should also be wirelessly connected to each other (Friedewald, et al., 2010; Mattern, 2005; Ramos, 2009). Consequently, computer services could be accessed at any place and at any time.

Relation to other terms. Its scientific origins and the focus on fundamental research distinguishes Ubiquitous Computing from more pragmatic approaches, such as the Internet of Things. This may be one reason why the term itself has hardly crossed the borders from computer science to other disciplines and it may also be the reason why it has not gained the same popularity as other terms (Radenkovic and Kocovic, 2017). Nonetheless, it has inspired computer science and anticipated many aspects of present-day technology. Some even argue that almost every concept of modern computer technology draws on the framework of omnipresence and invisibility (Friedewald, et al., 2010).

Ubiquitous Computing is frequently used synonymously with Pervasive Computing. While the former concept stems from Weiser, the latter was coined by IBM Research (Olson, et al., 2015; Want, 2009). Opinions differ on how both terms are related. Some argue that both refer to the same conceptual framework (Dhawan, et al., 2016; Mattern, 2005; Want, 2009). Others argue that the scientific community prefers Ubiquitous Computing. Meanwhile, economic actors tend to use Pervasive Computing (Mattern, 2005), which can be regarded as the pragmatic continuation of Ubiquitous Computing (Friedewald, et al., 2010; Olson, et al., 2015).

 

++++++++++

Internet of Things

Key question. How can ordinary objects be integrated into the Internet to generate additional value?

Core aspects. One of the most recent popular terms in the context of this paper is the Internet of Things [8] (Atzori, et al., 2010; Miorandi, et al., 2012; Ning, 2013; Olson, et al., 2015; Radenkovic and Kocovic, 2017). The Internet of Things was arguably coined by the founders of MIT’s Auto-ID Lab (https://autoid.mit.edu), especially by Kevin Ashton in 1999 and David L. Brooks in 2001 (Mukhopadhaya and Suryadewara, 2014; Radenkovic and Kocovic, 2017; Santucci, 2010). Atzori, et al. (2010) identified three main strands of research: (1) Internet-oriented visions focus on middleware enabling connectivity at any time, at any place, and of anything; (2) things-oriented visions focus on sensors, which connect the physical and the virtual world; and, (3) semantic-oriented visions focus on knowledge helping users to process a huge amount of accessible data. However, independent from this perspective, current visions of the Internet of Things are based on a common rationale: a network of physical and virtual objects, which communicate and cooperate with users, Internet services, and each other (Mattern and Flörkemeier, 2010). Access to the Internet is provided independently of place and time (Mukhopadhaya and Suryadewara, 2014). In detail, the term refers to the two terminological elements — Internet and things.

Internet. The primary concern of the Internet of Things is to connect physical and virtual objects (Loukides and Bruner, 2015; Miorandi, et al., 2012). This can be implemented at two levels: First, objects are wirelessly connected to each other or a hub; for example, a room-wide network of computers, furniture, or power outlets (Atzori, et al., 2010; Khodadadi, et al., 2017). In a second step, many small networks are connected to a bigger network; for example, several room-wide networks are combined to each other or to the Internet (Ning, 2013). Radenkovic and Kocovic (2017) refer to the Internet of Things as one of six areas of the Internet: Internet of Information, of Systems, of People, of Places, of Things, and of Virtual Identities (LeHong and Mahoney, 2012).

Things. The second aspect refers to everyday objects of the physical world (e.g., cutlery, furniture, or clothing), which serve as physical access points to the Internet (Atzori, et al., 2010) and generate additional value on three different levels. First, networked things ‘know’ more than ordinary objects because they can gather additional context information from their physical environment and from other areas of the Internet (Radenkovic and Kocovic, 2017). Second, networked things are, at the same time, part of the physical and the virtual world. They have a unique identifier and transmit their acquired environmental information onto the Internet, which is why the Internet of Things is often referred to as the unification of the physical world and the cyberworld (Friedewald, et al., 2010; Ning, 2013). Third, networked things can be made smart or intelligent more easily than unconnected objects.

Relation to other terms. The conceptual framework of the Internet of Things builds upon Ubiquitous Computing’s main idea of invisible and omnipresent technologies (Radenkovic and Kocovic, 2017) and enhances the aspect of connectivity by including the Internet. Initially, the Internet of Things was simply the extension of Ubiquitous Computing to the Internet (Olson, et al., 2015). Ubiquitous Computing implicitly refers to the integration of computer components into ordinary physical objects (Dhawan, et al., 2016; Mattern, 2005; Weiser and Brown, 1997) but this idea is more explicitly elaborated upon in the Internet of Things. Technological components are indeed physically invisible and may not even be recognised at all. Therefore, the Internet of Things is sometimes regarded as a complete implementation of Weiser’s concept (Mattern and Flörkemeier, 2010). Whereas Ubiquitous Computing is concerned with work-related interactions (e.g., replacing keyboards with touch or pen interfaces [Mattern, 2003]), the Internet of Things extends this idea to the everyday life and focuses on anthropomorphic interactions (e.g., speaking, seeing, hearing, or dealing with an object). The Internet of Things also provides a perfect infrastructure for smart or intelligent behavior and is closely connected to the concepts of Smart Objects (Atzori, et al., 2010; Miorandi, et al., 2012; Radenkovic and Kocovic, 2017; Olson, et al., 2015) and Artificial Intelligence (Mattern and Flörkemeier, 2010).

 

++++++++++

Smart Objects and Environments

Key question. How can objects and environments independently react to their users?

Core aspects. Unlike the other frameworks presented in this paper, the emergence of the term Smart Objects or Environments [9] cannot be attributed to a single person or group. Research on smart behavior of technology dates back to the late-1960s and early-1970s, although the meaning of smart was slightly different then and became more complex as technology developed (Cook, et al., 2009). By the standards of the 1960s, nearly everyone today lives in a smart home because the once futuristic thermostats or moving sensors have now become common.

The term smart implies the acquisition of context-information and the automated reaction to this context (Cook, et al., 2009; Kortuem, et al., 2010; Mattern, 2003; Kahlmann and Thalmann, 1999; Olson, et al., 2015). This requires the ability to sense information, and also requires processing power and at least basic connectivity (Gubbi, et al., 2013; Miorandi, et al., 2012; Olson, et al., 2015; Kortuem, et al., 2010). Reactions range from simple activity-aware recommendation of services and triggering of alarms up to complex process-aware reactions (Kortuem, et al., 2010).

Physical smart objects have a physical appearance and physical qualities (Miorandi, et al., 2012; Olson, et al., 2015). Research in this context focuses on coupling of resources, collective behavior (Augusto and Aghajan, 2009) and wireless communication among several objects (Gubbi, et al., 2013). Physical objects perceive their environment through multiple sensors (Augusto and Aghajan, 2009; Khodadadi, et al., 2017; Mattern, 2003; Olson, et al., 2015). Context is defined by other physical smart objects, ambient occurrences (e.g., location, light, sound, speech, or oxygen), and (predicted) needs of their users. Reactions range from interactions with the user to the physical manipulation of the environment (Augusto and Aghajan, 2009; Kortuem, et al., 2010; Mattern, 2003; Olson, et al., 2015).

Virtual smart objects imitate appearance and qualities of physical objects digitally. Research focuses on interactions between virtual objects and users or other virtual objects (Kahlmann and Thalmann, 1999). Interactions take place as reactions to a context, which is defined by other virtual objects and environments, or the user’s avatar. In addition to the features of physical smart objects, virtual objects are able to communicate their virtual interaction features, including intrinsic object attributes (e.g., movement or weight), interaction information (e.g., position of interaction components like door knobs or the virtual hand avatar of the user), and situation-specific object and agent behavior.

Relation to other terms. Ubiquitous Computing includes main issues of Smart Objects and Environments. Both conceptual frameworks feature a minimal set of communicative functions to accept and answer information in combination with computing capabilities (Miorandi, et al., 2012). While Ubiquitous Computing means adaption to contexts, Smart Objects and Environments extend this understanding to social interaction. Furthermore, Ubiquitous Computing is concerned with tools, disappearing from consciousness through adaption. In contrast, Smart Objects seem to be simple autonomous agents in the course of social interaction, making them visible in everyday contexts. The term smart became popular at the same time as the Internet of Things because both terms are closely linked to each other in research. (Atzori, et al., 2010; Gubbi, et al., 2013; Mattern and Flörkemeier, 2010; Miorandi, et al., 2012; Olson, et al., 2015; Radenkovic and Kocovic, 2017) Atzori, et al. (2010), for instance, classify research on Smart Objects as part of the ‘things’-oriented perspective in the Internet of Things. Meanwhile research on the Internet of Things concentrates on connection of things to the Internet, research on Smart Objects addresses the ability to act independently and context-sensitive. In other words, the Internet of Things establishes the framework for smart acting objects through connectivity. The origin of the term smart is closely connected to Artificial Intelligence (Radenkovic and Kocovic, 2017) because they occurred at a similar time. In addition, Ambient Intelligence (Cook, et al., 2009) is increasingly converging with research on Smart Environments. As we outlined previously, the understanding of smart has changed since its onset. The current development may be interpreted as further step towards a new understanding of smartness, growing together with intelligence.

 

++++++++++

Ambient Intelligence

Key question. How can environments autonomously and proactively support their inhabitants?

Core aspects. The term Ambient Intelligence [10] was coined in 1999 by Emile Aarts and picked up by the European Commission in 2001 (Friedewald, 2008; Cook, et al., 2009). Aarts (2004) envisioned an environment that recognises its inhabitants, adapts and responds to them, learns from their behavior and eventually expresses emotions. Furthermore, he referred to proactive interaction, which results in “enhanced efficiency, increased creativity and greater personal well-being” of humans [11]. Thus, Ambient Intelligence depends on adaptive and responsive computing technology embedded into ordinary environments (Cook, et al., 2009) and enriched by the key features of intelligence (Falomir, et al., 2017). It consists of two concepts — ambient and intelligent.

Ambient. The environment bases on connected, integrated, and implicit technology in the everyday physical surroundings (Ramos, 2009) with a focus on natural interaction (Aarts and Encarnação, 2006; Aarts and Wichert, 2009), equipped with sensor networks to collect a variety of physical information (Aarts and Wichert, 2009; Cook, et al., 2009).

Intelligence. Intelligence refers to the framework of Artificial Intelligence and the aim to benefit the environment’s inhabitant. Therefore, intelligence is related to context. Context models combine environmental attributes (Aarts, 2004; Cook, et al., 2009) with individual presence, behavioral patterns (Aarts and Encarnação, 2006; Aarts and Wichert, 2009), and cognitive and affective approaches of stress, attention, moods or needs (Falomir, et al., 2017; Ramos, 2009). Intelligent behavior also aims at social interaction (Aarts and Encarnação, 2006; Aarts, 2004; Aarts and Wichert, 2009). Consequently, the environment becomes a social agent and “the interaction of the user and the environment changes from unidirectional to bidirectional” [12]. Although decision-making and social interaction are explicit features of environments, humans and environments remain unequal interaction partners. Moreover, the aim is to design an agent that acts for the benefit of its inhabitants — like an “electronic butler[13] — adapts to them (Aarts and Wichert, 2009; Falomir, et al., 2017) and personalises its features (Ramos, 2009).

Relation to other terms. The conceptual framework of Ambient Intelligence is the result of implementing Artificial Intelligence into a physical environment, equipped with Ubiquitous Computing (e.g., Augusto and McCullagh, 2007; Cook, et al., 2009; Miorandi, et al., 2012; Ramos, 2009). The fundamental difference between Ubiquitous Computing and Ambient Intelligence is the purpose of computer technology in an environment. Ubiquitous Computing conceives of computer technology as an extension of human capabilities, similar to a tool (e.g. simple glasses [Weiser, 1994]), without being the focus of attention itself during utilisation. The interaction stays one-sided and the adaptive technology becomes invisible during using processes. In contrast, Ambient Intelligence aims to benefit its inhabitants — it recognises, adapts to and socially interacts with them. The environment becomes a ubiquitous social agent who is discretely visible during interaction, similar to a “butler” (Cook, et al., 2009; Olson, et al., 2015). Interaction becomes a two-sided social act. Ambient Intelligence emerged at the same time as the Internet of Things but has a different focus. Both share the intention to integrate computer technology into everyday objects and environments (Cook, et al., 2009). They differ in terms of the abilities of this integrated technology. While the Internet of Things specialises in communication and digital identity of objects and environments by connecting them to the Internet, Ambient Intelligence aims at creating environments as independent social actors. In other words, within the Internet of Things the Internet becomes ubiquitous in the environment, while within Ambient Intelligence the environment becomes an agent. There are only slight differences between the conceptions of Smart Environments and Ambient Intelligence, which may have facilitated the increasing synonymous use of these terms in research. However, the conceptions differ in the degree of agency, which we will outline later on.

 

++++++++++

Artificial Intelligence

Key question. How can computers think and act autonomously and proactively?

Core aspects. Although Alan Turing initially defined ‘intelligence of computer systems’ in 1950 (Barr and Feigenbaum, 1981; Ertel, 2016; Turing, 1950), the term Artificial Intelligence came into existence in 1956 at the Dartmouth Conference, when John McCarthy introduced the term and Allan Newell, Cliff Shaw, and Herbert A. Simon presented the first working AI-algorithm “The Logic Theorist” (Cohen and Feigenbaum, 1982; Ertel, 2016; Russell and Norvig, 2010). There have since been numerous studies on Artificial Intelligence, influenced by different theoretical and methodological approaches from various scientific disciplines (Russell and Norvig, 2010) [14]. Accordingly, it is the most scientific elaborated conceptual framework in the context of this paper, and it still is a “flourishing and rapidly developing field” (Cohen and Feigenbaum, 1982). A rough definition of Artificial Intelligence refers to autonomously perceiving, thinking and acting computer agents, which assist humans in difficult tasks, reactively by adapting to change and proactively by creating and pursuing goals (McCorduck, 2004; Russell and Norvig, 2010). Hence, computer technology is explicitly regarded as a social actor. Although there is a basic consensus, the specific definitions and conceptualisations vary greatly. The main differences concern the range of competence and the underlying model of intelligence (Russell and Norvig, 2010). Due to the limited space of this paper, we will focus on the former because it is more directly connected to the other frameworks presented in our paper. Competence ranges from problem-solving to general or holistic.

Problem-solving agents are restricted to certain tasks or situations, and based on highly specialised algorithms, which perform very successfully in their limited setting (Russell and Norvig, 2010). They serve as personal assistants for complicated tasks in predefined areas (McCorduck, 2004). Popular examples can be found in the areas of games (e.g., chess, Go or the multi-player game Quake III Arena), spam filtering, speech recognition, autonomous cars, or language translation.

General or holistic agents are able to perform in many different environments and situations. Consequently, they are more flexible and creative. Although the construction of these agents is more complicated, it does not seem impossible, especially in account of successful problem-solving agents (Russell and Norvig, 2010). Hence, holistic agents have recently been foregrounded by the terms Human-Level AI (HLAI) and Artificial General Intelligence (AGI). With increasing intelligence and range of competence, interventions by humans become less necessary. Social interaction becomes a main feature. Thus, objects and environments are active participants of real world processes and have an impact on social or economic life (Olson, et al., 2015; Santucci, 2010).

Relation to other terms. Artificial Intelligence and Ubiquitous Computing have contradictory intentions. While Ubiquitous Computing envisions computer technology as a tool that disappears from the consciousness during use, Artificial Intelligence seeks to turn computer technology into social agents. The proponents of Ubiquitous Computing even question the need of intelligence: “If a computer merely knows what room it is in, it can adapt its behavior in significant ways without requiring even a hint of artificial intelligence.” [15]. Although research on Artificial Intelligence considers technology integrated in environments and equipped with sensors as essential (Russell and Norvig, 2010), it does not focus on invisibility. Moreover, technological processes become visible through gestural, vocal or written two-sided interaction with its users. Which is why Weiser [16] explicitly challenges the implementation of intelligent agents: “Anything so insidiously appealing should immediately give pause. [...] A computer I need to talk to, give commands to, or have a relationship with [...], is a computer that is too much the center of attention.” As outlined earlier, the Internet of Things provides the infrastructure for intelligent behavior through the connection of objects to the Internet (Russell and Norvig, 2010). Therefore, research on the Internet of Things increasingly focuses on ‘intelligent’ instead of merely ‘smart’ objects (e.g., Friedewald, et al., 2010; Mukhopadhaya and Suryadewara, 2014; Santucci, 2010). While Artificial Intelligence aims at modeling or simulating humanlike or rational proactive thinking and behavior (Russell and Norvig, 2010) research on Smart Objects has a more narrow focus on adaptive and appropriate reactions in relation to an environment. The conceptual frameworks of Artificial Intelligence and Ambient Intelligence share the understanding of computer technology as autonomous and proactive agents. In contrast to Artificial Intelligence, Ambient Intelligence consequently integrates these agents into a physical environment, which inhabits people (Cook, et al., 2009; Falomir, et al., 2017). Human or animal robots, as well as virtual intelligences are not the focus of attention.

 

++++++++++

Disentangling the terminology

Although all conceptual frameworks are highly interrelated, it appears fruitful to semantically disentangle them for analytical purposes (see Figure 1). In particular, Ubiquitous Computing and Artificial Intelligence can be regarded as the key conceptions of technological progress during the last 30 years. The other conceptual frameworks in the context of this paper range between these two.

 

Interrelatedness and dimensions of the conceptual frameworks
 
Figure 1: Interrelatedness and dimensions of the conceptual frameworks.
 
Note: Larger version of Figure 1 available here.

 

On the one hand, Ubiquitous Computing understands computer technology as a tool and is concerned with its integration into work contexts, processes, and environments. The aim of this approach is to recede technological components into the background. Consequently, their appearance and operations become invisible and omnipresent. The intention is to extend human capabilities with the help of tools, which compensate human weaknesses, similar to glasses, helping us to see the world properly without attracting attention to themselves (Weiser, 1994).

On the other hand, Artificial Intelligence understands computer technology as discrete entity, which acts as an efficient and multifunctional agent. The conception of this agent bases on models of human (or reasonable) thinking and behavior and the normative approach of ‘benefit humans’. Computer technology becomes an ‘electronic butler’, who acts independently and interacts socially with its users. Therefore, it becomes an explicit partner of social interaction.

The Internet of Things and Ambient Intelligence adopt the idea of integration from Ubiquitous Computing and extend it from work contexts to universal contexts. However, they differ in the way they conceptualise these everyday environments: the Internet of Things focuses on connecting environments to a worldwide Internet, while Ambient Intelligence emphasises their intelligent behavior. The Internet of Things constitutes a framework for Smart Objects, making them more efficient due to the connection to the Internet. Smart Objects and Environments constitutes a conceptual bridge. On the one hand, the term refers to embedding networked computer technology into the physical environment of people, which connects to Ubiquitous Computing and Internet of Things. On the other hand, it includes the capability of simple independent interaction, relating to the conceptual frameworks of Ambient and Artificial Intelligence. Smart Objects and Artificial Intelligence share the focus on technological assistance. The main difference between the approaches concerns the level of capability: interaction with Smart Objects remains on the level of automated reaction to collected data and context information, while intelligent technologies include autonomous proactive behavior.

 

++++++++++

Extraction of underlying dimensions

Drawing on Friedewald, et al. (2010), Siemoneit (2003), Mattern and Flörkemeier (2010) we have identified four dimensions underlying the above presented concepts: connectivity, invisibility, awareness, and agency. These dimensions can be assigned to two super-dimensions — connectivity and invisibility deal with aspects of integration, while awareness and agency are concerned with intelligence issues. Integration focuses on natural interaction between humans and computers, which is accomplished through invisible technological components and wireless connection. Consequently, human-computer interaction becomes omnipresent, intuitive, and less apparent. In this paper, intelligence aims to enrich interaction. The ability to be aware of the environment is required to perform different levels of agency, ranging from simple tasks to very complex social behavior. Although the dimensions overlap conceptually, we separate them for analytical reasons (see Tables 1 and 2). Consequently, social implications of the current technological changes can be addressed more precisely. We will next describe these dimensions in more detail and illustrate specific social implications.

 

Table 1: Overview of the conceptual frameworks
Ubiquitous ComputingInternet of ThingsSmart Objects/ EnvironmentsAmbient IntelligenceArtificial Intelligence
SynonymsPervasive Computing, Everyday Computing, EverywareInternet of Everything, Web of Things, Real World Internet, Ubiquitous WebResponsive Objects/Environments, Smart Items, Smart SpimeIntelligent Environment, Intelligent Object --
Key QuestionsHow can computers blend in the natural human mode of dealing with tools?How can ordinary objects be integrated in the Internet to generate additional value?How can objects and environments automatically react to their users?How can environments autonomously and proactively support their inhabitants?How can computers think and act autonomously and proactively?
Popular ExamplesSmartphone, Pad, Smart-TV, Smart-Board Smart GlassesSmart Home/City Smart Fridge/Light/Heating/Vacuum Cleaner Smart Loudspeaker (e.g., Amazon Echo, Apple Homepod, Essentials Home, ...)Self-Driving Car, Intelligent HomesIBM Deep Blue, Google Alpha Go, Open AI Five, Google Brain with Google Translate, Google Voice or Google Images, Microsoft Project Adam, Facebook with Textanalysis, Facial Recognition, Twitter Timeline, Watson IBM Conversational assistants using natural language (e.g., Amazon Alexa, Apple Siri, Essential Ambient OS, Microsoft Cortana, Samsung Bixby)

 

 

Table 2: Conceptual frameworks and underlying dimensions
Ubiquitous ComputingInternet of ThingsSmart Objects/ EnvironmentsAmbient IntelligenceArtificial Intelligence
ConnectivityWireless connection of computer devicesWireless connection of ordinary objects to the InternetWireless connection of ordinary objects among each other
Transferring data and equipping connected objects with additional information
 
InvisibilityPhysicalTechnology is physically invisible because it resembles ordinary objects. It blends in the natural mode of using tools (e.g.,. glasses)Technology is physically invisible due to its integration in ordinary objects or physical environments. Therefore, it blends in the natural human manipulation modes (e.g., turning, handling, touching)Technology is physically invisible in some fields (e.g., cybernetics) due to support of human manipulation modes.
OperationalTechnology is operationally invisible during utilization.Technology is operationally visible due to (human) communication modes (e.g.,, blinking, beeping, speaking, answering, body language)
 
AwarenessSensation of attributes and behavior of environment and users, related to work taskSimple sensation of specific physical and virtual attributes and behavior of environment and usersComplex sensation of physical and virtual attributes and behavior of environments and users
 
AgencyTechnology adapts during utilizationReactive interaction, automated (and simple autonomous) processes, manipulation of the physical environmentProactive interaction, autonomous assistant ('electronic butler), manipulation of the physical environmentProactive interaction, autonomous partner

 

Connectivity

Connectivity relates to (wireless or wired) networks of different devices, objects, or other networks aiming at sending and receiving data, to communicate, cooperate, or extend capabilities. Important technological features of this dimension are unique identification and addressability of connected elements, which enables the possibility of adding further information and bandwidth (Mattern, 2003; Mattern and Flörkemeier, 2010; Siemoneit, 2003). Wireless connections entail further attributes such as different mobile distances (e.g., Body, Personal, Local and Wide Area Networks), as well as the durability of networks, e.g., static in contrast to Ad Hoc Networks, which emerge spontaneously and are self-managed between objects or devices, independent from a local base (Friedewald, et al., 2010; Siemoneit, 2003; Zhang, et al., 2017).

Relation to terms. All of the terms presented in this paper depend on connectivity as basic infrastructure. Ubiquitous Computing focuses on the connection of devices at the level of Local Area Networks. Smart Objects and Internet of Things transfer connectivity to ordinary objects and expand the connection to the global level of Wide Area Networks. Ad Hoc Networks and the identification of connected objects are highly relevant under the term Internet of Things, while Smart Objects, Ambient, and Artificial Intelligence are concerned with transferring data and equipping connected objects with additional resources such as information and abilities.

Social implications. The mere number of interconnected devices leads to an increased network of devices and their human owners, resulting in high-ranged social networks. Global networking of this magnitude increases social relationship and relatedness to a worldwide scale. Relevant to this are emerging publics, as well as activation, use, and consequences of weak and strong ties at the level of vast networks. The transformation of the wired network to wireless networking supports a perpetual connection, providing unlimited local and temporal access to services and social ties. This development is referred to as “always on” (Santucci, 2010; Turkle, 2008) or “permanently connected” (Vorderer and Kohring, 2013; Vorderer, et al., 2017). The emphasis in this context anticipates issues that are related to modified patterns and conditions of usage. Furthermore, empirical studies have demonstrated that the mere presence of cell phones and social media may reduce human task performance (Thornton, et al., 2014; Toma, 2013; Ward, et al., 2017). Consequently, the competence concerning concentration or conversation quality may be challenged in this respect. Ad Hoc Networks, which evolve or dissolve as required while remaining mostly independent of static lines, influence social and economic conditions due to increased decentralisation, the flexibility of access, and the replacement of former monopolies of infrastructure. This may, for instance, affect the query of digital inequality between rural and urban regions, and between industrial and developing countries (DiMaggio, et al., 2001; Scheerder, et al., 2017). The stability and the continuously increasing bandwidth of the networks support the external storage of digitised content, which thus can be accessed if required (by using services such as Spotify, Wikipedia, or Netflix). Thus, the physical ownership of classical (storage) media such as books, CDs, videos or cassettes is changing from a necessity of receiving media content into a voluntary decision, which in turn implicates a change in the motives of such ownership.

Invisibility

Invisibility refers to the degree to which humans notice computer technology while interacting with it, which is increased by reducing the technologys perceived complexity and artificiality (Friedewald, et al., 2010; Mattern and Flörkemeier, 2010; Siemoneit, 2003). This dimension can be divided into two subdimensions: operational and physical invisibility. On an operational level, the technology becomes invisible through habitualisation (Robben and Schelhowe, 2012). A prominent example is the integration of smartphones into everyday life. In addition, technology may pass unnoticed if it acts silently and is hidden in the background through adapting to specific situations. Another way of letting the technology disappear is to use human manipulation modes during interaction. In other words, interfaces are anthropomorphised and intuitive interaction modes of manipulation (e.g., lifting, moving, turning, and touching) are supported (Hellige, 2008; Friedewald, 2008). Physical invisibility refers to the integration of computer technology into ordinary objects, such as plant pots, light bulbs, or heaters (Friedewald, et al., 2010; Mattern and Flörkemeie, 2010; Siemoneit, 2003). A milder approach is making technology resemble ordinary objects (e.g., e-books or digital pens).

Relation to terms. The conceptual framework of Ubiquitous Computing addresses all facets of operational invisibility, except communicative interaction. Physical invisibility is mainly achieved through resembling ordinary objects; that is, offering different sizes, shapes or surfaces of computer devices for different purposes, such as smartphones, computer tablets, electronic whiteboards, or e-books. The Internet of Things, Smart Objects, and Ambient Intelligence focus on physical invisibility by embedding technological components into ordinary objects or environments. They achieve operational invisibility through adaptive behavior, human manipulation modes and embedding technology into everyday practice. Due to the focus of the latter two conceptions on modes of human communication, they are visible during interaction. Artificial Intelligence does not explicitly refer to physical invisibility but it does include operational invisibility through adaptive behavior and, at the same time, visibility through its focus on assistance and human communication.

Social implications. This dimension concerns the perception of the existence and performance of technologies by its users. The physical integration of technological components into a wide range of everyday objects makes digital capabilities omnipresent. Consequently, the complexity and capacities of the devices increase but the perceived artificiality decreases. Digitalised environments are indistinguishable from natural ones. On the operational level, the digitised daily objects ‘work’ and ‘behave’ as discreetly as possible and embed themselves in the everyday actions of the user by performing simple adaptions and operations in the background. Therefore, users only receive the information and services needed for the task, enabling them to concentrate on the task, which means that our competence should be positively affected (Friedewald, et al., 2010; Want, 2009). Especially in this context, cultural and social influences on the recognition of affordances and competence become relevant due to invisibility. What is unclear, however, is what kind of competence is promoted if we can use the technologies more efficiently but hardly understand when and how they work. Issues of critical assessment of delivered information, which require knowledge of specific processes, are located in this dimension. Consequently, origins, affordances, effects, risks or problems of technological processes may be overestimated or underestimated.

Awareness

Awareness refers to the ability of devices or objects to collect and store current and long-term data about their own and their user’s situation (Friedewald, et al., 2010; Mattern and Flörkemeier, 2010). The collected information ranges from simple to complex awareness (Abowd, et al., 1999). The data originates in the physical world (e.g., location, air quality, ambient light, or sound), as perceived through different sensor technologies, which can be classified in sensors with external power supply and wireless sensors (Friedewald, et al., 2010). If devices or objects are connected to a network, further data can be collected in the virtual world. Virtual data includes additional sensor data of other objects, previous uses or the user’s behavior in other contexts (Abowd, et al., 1999). The more sensors are involved, the more often they are activated and the bigger the connected network, the more complex perceived contexts become and the higher the awareness of the device or object can be considered.

Relation to terms. This dimension is reflected well in the presented frameworks but serves different purposes. Ubiquitous Computing focuses on work-related perceptions. As the Internet of Things and Smart Objects mainly aim at mapping the physical everyday world and at reacting to it, their concept of awareness is limited to specific situations. Ambient Intelligence aims at the perception of the physical environment but shares a complex understanding of context with Artificial Intelligence. Artificial Intelligence requires a high level of awareness to improve its reactions through learning, as well as to perform proactive behavior.

Social implications. Awareness concerns the interplay of being aware of the environmental situation of the technology and self-awareness of its position in this environment. The awareness of the environment refers to the presence and behavior of objects and subjects in a particular context. In addition to collecting emerging events, it derives laws, goals, and motives from patterns to prognosticate expected environmental situations. Therefore, collecting information is inseparably linked to storing, sharing, and linking data. The fundamental conflict between the functionality of the service and the endangered privacy of the user is mainly relevant to this issue. The quantity of collected data contains the potential for more complex and extended analysis, which, for example, may increase comfort. However, the meaningfulness of extensive data collection is still disputable (see discussions under the keyword Big Data). The various individual and social consequences of data access and ownership range between surveillance by others and self-monitoring.

A slightly different question is the self-awareness of technology. It refers to the comprehension of one’s position relative to the objects and subjects of the environment, an ability that up to now has been reserved for human beings and a few animals defined as intelligent, and which is associated, for example, with the development of identity.

Agency

We have identified agency as the fourth dimension of modern computer technology. Although the conceptual frameworks vary widely in terms of the substance of capacity, they consensually refer to it as the ability of devices or objects to independently use collected data to initiate activities or interactions, which have an impact on the physical or virtual world. Hence, the independent behavior is crucial to all previously depicted frameworks but ranges from automation to autonomy (Hancock, 2017) and from reactive to proactive behavior. Linked to this dimension are the terms smartness and intelligence, which are frequently used interchangeably and which still lack a standard definition (Hernández and Reiff-Marganiec, 2014). Nevertheless, according to the conceptual frameworks that we have mentioned, an enhanced ability to combine, reason and embed information in a broad context improves the agency of the technology. Depending on the degrees of this contextual reasoning, in addition to increased proactive behavior and extended autonomy, an object’s capability evolves from automation to smart behavior to intelligent agency.

Relation to terms. Context-adaption and smartness of technology without being intelligent is a core idea of Ubiquitous Computing. Any kind of intelligence is rejected on account of their visibility. Whereas, the Internet of Things frequently explores automated, global networked behavior, Smart Objects is concerned with an increased ability to combine and use collected data to react to the environment. Ambient Intelligence and Artificial Intelligence emphasise extensive contextual reasoning and increased autonomous and proactive behavior. However, the boundaries are becoming indistinct.

Social implications. Because the impact of technology is particularly evident in actions, this domain of research is heavily investigated to the present day. Consequently, we will only outline a few critical questions. While the dimension of awareness aims at the scope, quality, and relation of environmental information, this knowledge constitutes the framework for action in the dimension of agency. Technology as an actor becomes relevant. If, for example, activities refer directly to other humans, then aspects of social interaction and relatedness become apparent. Turkle (2011) has argued that intensifying relations to machines has alienated us from other humans and that individuals increasingly prefer computer-mediated relations to intrapersonal relations. Competence, concerning task performance, is equally relevant. Research has shown that the risk of adverse effects on human performance rises with the degree of automation (Onnasch, et al., 2014). Furthermore, the degree of the autonomous behavior brings the norms and values underlying these actions to the center. In addition, the self-governed actions of the technology may be related to the perceived autonomy of the user during interaction. Reactance theory (Brehm and Brehm, 1981) postulates that individuals react negatively toward perceived threats to autonomy. Although machine agency may have the potential to strengthen human agency (Følstad, et al., 2017), empirical research has demonstrated that machine agency is also frequently perceived as a threat (Rijsdijk and Hultink, 2003; Roubroeks, et al., 2011; Schweitzer and Van den Hende, 2016). Some scholars have even proposed to restrain machine agency for the sake of human autonomy (Hancock, 2017).

 

++++++++++

Conclusion

This paper presents the most relevant concepts used to describe current technological applications in everyday life and links them to each other. It described the progressive merging of invisible tools, working in the background (Ubiquitous Computing), integrated into every domain of life and connected to the Internet (Internet of Things), with independently thinking and acting actors (Artificial Intelligence), implemented both into environments (Ambient Intelligence), and into everyday objects (Smart Objects). The challenge of the imprecise definition of concepts and the associated ambiguous use in research was addressed by elaborating the most critical questions and key concepts behind the terminology and by demonstrating the differences between them (see Table 1). Although the concepts have vivid and broad visions about the integration of technology in everyday situations, their blurred and partly overlapping boundaries limit their suitability for empirical sociological research.

We extracted the four dimensions of modern technologies: Connectivity, Invisibility, Awareness, and Agency. Contrary to the terminology that has been described, these dimensions are distinct from each other and point to severe social issues. Hence, concrete conclusions about their social impact can be drawn and linked more easily to previous theoretical findings — transcending the boundaries of individual technologies (see Table 2). A further benefit is the visibility of gaps in extant research. While some dimensions, such as Connectivity or Agency, have already received a considerable amount of attention, Invisibility has not yet been taken into serious consideration.

Operationalisability is a significant advantage of the proposed dimensions, which is why they are particularly interesting for empirical social research. As described in this paper, several concepts — as well as related technology applications — address these dimensions to different degrees. Thus enabling, for example, the interpretation of conflicting results about the same technology.

Over the next few years, we expect continued significant technological and related societal changes. In this context, social science must continually face the challenge to ensure that its results remain valid beyond a handful of years. To mitigate the race against emerging developments, our article provides a tool to examine socio-cultural changes strategically and systematically while avoiding dependence on particular technology developments or visions. Therefore, we aim to facilitate the foresighted and more active participation of social science research in this process. End of article

 

About the authors

Katrin Etzrodt is a Ph.D. student at the Technische Universität Dresden in the Institute of Communication and Media. She gratefully acknowledges the support and generosity of the Scholarship Program for the Promotion of Early-Career Female Scientists of TU Dresden, without which the present paper would not have been attainable.
E-mail: katrin [dot] etzrodt [at] tu-dresden [dot] de

Sven Engesser is a professor in the Institute of Communication and Media at the Technische Universität Dresden.
E-mail: sven [dot] engesser [at] tu-dresden [dot] de

 

Notes

1. Santucci, 2010, p. 12.

2. Other terms referring to a similar concept but less popular are: Pervasive Computing, Nomadic Computing, Everyday Computing, and Everyware.

3. Weiser, 1993, p. 75.

4. Weiser, 1991, p. 98.

5. Weiser, 1991, p. 94.

6. Weiser, 1991, p. 98.

7. Ibid.

8. Other terms referring to a similar concept but less popular are: Internet of Everything, Web of Things, Real World Internet, and Ubiquitous Web.

9. Other terms referring to a similar concept but less popular are: Intelligent Objects/Environments, Responsive Objects/Environments, and Smart Items.

10. Other terms referring to a similar concept but less popular are: Intelligent Environments and Intelligent Objects.

11. Aarts and Wichert, 2009, p. 244.

12. Aarts and Encarnação, 2006, p. 11.

13. Cook, et al., 2009, p. 278.

14. The works of Stuart Russell and Peter Norvig (2010), Wolfgang Ertel (2016), and Pamela McCorduck (2004) provide elaborated overviews of the history of the term and research on Artificial Intelligence.

15. Weiser, 1991, p. 99.

16. Weiser, 1994, p. 7.

 

References

E. Aarts, 2004. “Ambient intelligence: A multimedia perspective,” IEEE Multimedia, volume 11, number 4, pp. 12–19.
doi: https://doi.org/10.1109/MMUL.2004.1261101, accessed 15 August 2019.

E. Aarts and R. Wichert, 2009. “Ambient intelligence,” In: H.-J. Bullinger (editor). Technology guide: Principles — applications — trends. Berlin: Springer-Verlag, pp. 244–249.
https://doi.org/10.1007/978-3-540-88546-7_47, accessed 15 August 2019.

E. Aarts and J. Encarnação (editors), 2006. True visions: The emergence of ambient intelligence. Berlin: Springer-Verlag.
https://doi.org/10.1007/978-3-540-28974-6, accessed 15 August 2019.

G.D. Abowd, A.K. Dey, P.J. Brown, N. Davies, M. Smith, and P. Steggles, 1999. “Towards a better understanding of context and context-awareness,” In: H.W. Gellersen (editor). Handheld and ubiquitous computing. Lecture Notes in Computer Science, volume 1707. Berlin: Springer, pp 304–307.
https://doi.org/10.1007/3-540-48157-5_29, accessed 15 August 2019.

L. Atzori, A. Iera, and G. Morabito, 2010. “The Internet of things: A survey,” Computer Networksz, volume 54, number 15, pp. 2,787–2,805.
https://doi.org/10.1016/j.comnet.2010.05.010, accessed 15 August 2019.

J.C. Augusto and H. Aghajan, 2009. “Editorial: Inaugural issue,” Journal of Ambient Intelligence and Smart Environments, volume 1, number 1, pp. 1–4.
https://doi.org/10.3233/AIS-2009-0011, accessed 15 August 2019.

J.C. Augusto and P. McCullagh, 2007. “Ambient intelligence: Concepts and applications,” Computer Science and Information Systems, volume 4, number 1, pp. 1–27.
https://doi.org/10.2298/CSIS0701001A, accessed 15 August 2019.

A. Barr and E.A. Feigenbaum (editors), 1981. The handbook of artificial intelligence, volume 1. Stanford, Calif.: HeurisTech Press; version at https://archive.org/details/handbookofartif01barr, accessed 15 August 2019.

S.S. Brehm and J.W. Brehm, 1981. Psychological reactance: A theory of freedom and control. New York: Academic Press.

P.R. Cohen and E.A. Feigenbaum (editors), 1982. The handbook of artificial intelligence, volume 3. Stanford, Calif.: HeurisTech Press; version at https://archive.org/details/handbookofartific03cohe, accessed 15 August 2019.

D.J. Cook, J.C. Augusto, and V.R. Jakkula, 2009. “Ambient intelligence: Technologies, applications, and opportunities,” Pervasive and Mobile Computing, volume 5, npp. 277–298.
https://doi.org/10.1016/j.pmcj.2009.04.001, accessed 15 August 2019.

S.M. Dhawan, B.M. Gupta, and R. Gupta, 2016. “Global pervasive and ubiquitous computing during 2005–14,” Annals of Library and Information Studies, volume 63, number 2, at http://op.niscair.res.in/index.php/ALIS/article/view/11391, accessed 15 August 2019.

W. Ertel, 2016. Grundkurs Künstliche Intelligenz: Eine praxisorientierte Einführung. Fourth revised edition. Wiesbaden: Springer Fachmedien.
https://doi.org/10.1007/978-3-8348-9441-0, accessed 15 August 2019.

Z. Falomir, J.A. Ortega, N. Martinez, and H. Guesguen, 2017. “Thematic issue on human-centred ambient intelligence: Cognitive approaches, reasoning and learning,” Journal of Ambient Intelligence and Smart Environments, volume 9, number 5, pp. 539–542.
https://doi.org/10.3233/AIS-170449, accessed 15 August 2019.

A. Følstad, V. Engen, I.M. Haugstveit, and J.B. Pickering, 2017. “Automation in human-machine networks: How increasing machine agency affects human agency,” arXiv (24 February), at https://arxiv.org/abs/1702.07480, accessed 15 August 2019.

M. Friedewald, 2008. “Ubiquitous Computing: Ein neues Konzept der Mensch-Computer-Interaktion und seine Folgen,” In: H.D. Hellige (editor). Mensch-Computer-Interface. Zur Geschichte und Zukunft der Computerbedienung. Bielefeld: transcript, pp. 259–280.
https://doi.org/10.14361/9783839405642-008, accessed 15 August 2019.

M. Friedewald, O. Raabe, P. Georgieff, D.J. Koch, and P. Neuhusler, 2010. Ubiquitäres Computing: Das “Internet der Dinge” — Grundlagen, Anwendungen, Folgen. Studien des Büros fur Technikfolgen-Abschätzung beim Deutschen Bundestag, volume 31. Berlin: Edition sigma.
https://doi.org/10.5771/9783845269955-1, accessed 15 August 2019.

S. Greengard, 2015. The Internet of Things. Cambridge, Mass.: MIT Press.

J. Gubbi, R. Buyya, S. Marusic, and M. Palaniswami, 2013. “Internet of Things (IoT): A vision, architectural elements, and future directions,” Future Generation Computer Systems, volume 29, number 7, pp. 1,645–1,660.
https://doi.org/10.1016/j.future.2013.01.010, accessed 15 August 2019.

P.A. Hancock, 2017. “Imposing limits on autonomous systems,” Ergonomics, volume 60, number 2, pp. 284–291.
https://doi.org/10.1080/00140139.2016.1190035, accessed 15 August 2019.

H.D.Hellige, 2008. “Krisen- und Innovationsphasen in der Mensch-Computer-Interaktion,” In: H.D. Hellige (editor). Mensch-Computer-Interface. Zur Geschichte und Zukunft der Computerbedienung. Bielefeld: transcript, pp. 11–92.
https://doi.org/10.14361/9783839405642-001, accessed 15 August 2019.

M.E.P. Hernández and S. Reiff-Marganiec, 2014. “Classifying smart objects using capabilities,” 2014 International Conference on Smart Computing.
https://doi.org/10.1109/SMARTCOMP.2014.7043873, accessed 15 August 2019.

M. Kahlmann and D. Thalmann, 1999. “Modeling objects for interaction tasks,” In: B. Arnaldi and G. Hégron (editors). Computer animation and simulation ’98. Vienna: Springer, pp. 73–86.
https://doi.org/10.1007/978-3-7091-6375-7_6, accessed 15 August 2019.

F. Khodadadi, A.V. Dastjerdi, and R. Buyya, 2017. “Internet of Things: an overview,” arXiv (19 March), at https://arxiv.org/abs/1703.06409, accessed 15 August 2019.

G. Kortuem, F. Kawsar, V. Sundramoorthy, and D. Fitton, 2010. “Smart objects as building blocks for the Internet of Things,” IEEE Internet Computing, volume 14, number 1, pp. 44–51.
https://doi.org/10.1109/MIC.2009.143, accessed 15 August 2019.

J. Lave and E. Wenger, 1991. Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press.

H. LeHong and J. Mahoney, 2012. “Innovation insight: The ‘Internet of everything’ innovation will transform business,” Gartner Research (3 January), at https://www.gartner.com/doc/1886915/innovation-insight-internet-everything-innovation, accessed 15 August 2019.

M. Loukides and J. Bruner, 2015. “What is the Internet of Things?” O’Reilly Media (15 January), at https://www.oreilly.com/ideas/what-is-the-internet-of-things, accessed 15 August 2019.

F. Mattern, 2005. “Pervasive/Ubiquitous Computing,” Gesellschaft für Informatik (28 July), at https://gi.de/service/informatiklexikon/detail/pervasiveubiquitous-computing/, accessed 15 August 2019.

F. Mattern, 2003. “Vom Verschwinden des Computers — Die Vision des Ubiquitous Computing,” In F. Mattern (editor). Total vernetzt: Szenarien einer informatisierten Welt. Berlin: Springer-Verlag, pp. 1–41.
https://doi.org/10.1007/978-3-642-55550-3_1, accessed 15 August 2019.

F. Mattern and C. Flörkemeier, 2010. “Vom Internet der Computer zum Internet der Dinge,” Informatik Spektrum, volume 33, number 2, pp. 107–'121.
https://doi.org/10.1007/s00287-010-0417-7, accessed 15 August 2019.

P. McCorduck, 2004. Machines who think: A personal inquiry into the history and prospects of artificial intelligence. Twenty-fifth anniversary update. Natick, Mass.: A.K. Peters.

D. Miorandi, S. Sicari, F. De Pellegrini, and I. Chlamtac, 2012. “Internet of things: Vision, applications and research challenges,” Ad Hoc Networks, volume 10, number 7, pp. 1,497–1,516.
https://doi.org/10.1016/j.adhoc.2012.02.016, accessed 15 August 2019.

S.C. Mukhopadhaya and N. Suryadewara, 2014. “Internet of things: Challenges and opportunities,” In: S.C. Mukhopadhaya (editor). Internet of things: Challenges and opportunities. Cham, Switzerland: Springer, pp. 1–17.
https://doi.org/10.1007/978-3-319-04223-7_1, accessed 15 August 2019.

H. Ning, 2013. Unit and ubiquitous Internet of things. Boca Raton, Fla.: CRC Press.
https://doi.org/10.1201/b14742, accessed 15 August 2019.

N. Olson, J.M. Nolin, and G. Nelhans, 2015. “Semantic web, ubiquitous computing, or Internet of things? A macro-analysis of scholarly publications,” Journal of Documentation, volume 71, number 5, pp. 884–916.
https://doi.org/10.1108/JD-03-2013-0033, accessed 15 August 2019.

L. Onnasch, C.D. Wickens, H. Li, and D. Manzey, 2014. “Human performance consequences of stages and levels of automation: An integrated meta-analysis,” Human Factors, volume 56, number 3, pp. 476–488.
https://doi.org/10.1177/0018720813501549, accessed 15 August 2019.

B. Radenkovic and P. Kocovic, 2017. “From ubiquitous computing to the Internet of things,” In: P. Kocovic, R. Behringer, M. Ramachandran, and R. Mihajlovic (editors). Emerging trends and applications of the Internet of things. Hershey, Pa.: IGI Global, pp. 1–42.
https://doi.org/10.4018/978-1-5225-2437-3.ch001, accessed 15 August 2019.

C. Ramos, 2009. “Ambient intelligence environments,” In: J.R.R. Dopico, J. Dorado, and A. Pazos (editors). Encyclopedia of artificial intelligence. Hershey, Pa.: IGI Global, pp. 92–98.
https://doi.org/10.4018/978-1-59904-849-9.ch014, accessed 15 August 2019.

S.A. Rijsdijk and E.J. Hultink, 2003. “‘Honey, have you seen our hamster?’ Consumer evaluations of autonomous domestic products, Journal of Product Innovation Management, volume 20, number 3, pp. 204–216.
https://doi.org/10.1111/1540-5885.2003003, accessed 15 August 2019.

B. Robben and H. Schelhowe, 2012. “Was heißt be-greifbare Interaktion,” In: B. Robben and H. Schelhowe (editors). Be-greifbare Interaktion: Der allgegenwrtige Computer: Touchscreens, Wearables, Tangibles und Ubiquitous Computing. Bielefeld: transcript, pp. 7–16.
https://doi.org/10.14361/transcript.9783839420058.41, accessed 15 August 2019.

M. Roubroeks, J. Ham, and C. Midden, 2011. “When artificial social agents try to persuade people: The role of social agency on the occurrence of psychological reactance,” International Journal of Social Robotics, volume 3, number 2, pp. 155–165.
https://doi.org/10.1007/s12369-010-0088-1, accessed 15 August 2019.

S. Russell and P. Norvig, 2010. Artificial intelligence: A modern approach. Third edition. Boston: Pearson.

G. Santucci, 2010. “The Internet of things: Between the revolution of the Internet and the metamorphosis of objects,” In: P. Friess, P. Guillemin, and H. Sundmaeker, P. Guillemin, P. Friess, and S. Woelfflée; (editors). Vision and challenges for realising the Internet of things. Brussels: European Commission, pp. 11–24, and at http://www.internet-of-things-research.eu/pdf/IoT_Clusterbook_March_2010.pdf, accessed 11 November 2018.

F. Schweitzer and E. Van den Hende, 2016. “To be or not to be in thrall to the march of smart products,” Psychology & Marketing, volumee 33, number 10, pp. 830–842.
https://doi.org/10.1002/mar.20920, accessed 15 August 2019.

G. Siemoneit, 2003. “Ubiquitous Computing. Neue Dimensionen technischer Kultur,” TRANS Internet-Zeitschrift für Kulturwissenschaften, number 15, at http://inst.at/trans/15Nr/10_4/siemoneit_oliver15.pdf, accessed 15 August 2019.

L.A. Suchman, 1985. “Plans and situated actions: The problem of human-machine communication,” Xerox PARC Technical Report, ISL6, Corporate accession P85-0000S.

B. Thornton, A. Faires, M. Robbins, and E. Rollins, 2014. “The mere presence of a cell phone may be distracting,” Social Psychology, volume 45, number 6, pp. 479–488.
https://doi.org/10.1027/1864-9335/a000216, accessed 15 August 2019.

C.L. Toma, 2013. “Feeling better but doing worse: Effects of Facebook self-presentation on implicit self-esteem and cognitive task performance,” Media Psychology, volume 16, number 2, pp. 199–220.
https://doi.org/10.1080/15213269.2012.762189, accessed 15 August 2019.

A.M. Turing, 1950. “Computing machinery and intelligence,” Mind, volume 59, number 236, pp. 433–460.
https://doi.org/10.1093/mind/LIX.236.433, accessed 15 August 2019.

S. Turkle, 2011. Alone together: Why we expect more from technology and less from each other. New York: Basic Books.

S. Turkle, 2008. Always-on/Always-on-you: The tethered self. In: J.E. Katz (editor). Handbook of mobile communication studies. Cambridge, Mass.: MIT Press, pp. 121–137.
https://doi.org/10.7551/mitpress/9780262113120.003.0010, accessed 15 August 2019.

P. Vorderer and M. Kohring, 2013. “Permanently online: A challenge for media and communication research,” International Journal of Communication, volume 7, pp. 188–196, and at https://ijoc.org/index.php/ijoc/article/view/1963, accessed 15 August 2019.

P. Vorderer, D. Hefner, L. Reinecke, and C. Klimmt (editors), 2017. Permanently online, permanently connected: Living and communicating in a POPC world. New York: Routledge.
https://doi.org/10.4324/9781315276472, accessed 15 August 2019.

R. Want, 2009. “An introduction to ubiquitous computing,” In: J. Krumm (editor). Ubiquitous computing fundamentals. Boca Raton, Fla.: Chapman & Hall/CRC Press, pp. 1–35.

A.F. Ward, K. Duke, A. Gneezy, and M.W. Bos, 2017. “Brain drain: The mere presence of one’s own smartphone reduces available cognitive capacity,” Journal of the Association for Consumer Research, volume 2, number 2, pp. 140–154.
https://doi.org/10.1086/691462, accessed 15 August 2019.

M. Weiser, 1994. “The world is not a desktop,” Interactions, volume 1, number 1, pp. 7–8.
https://doi.org/10.1145/174800.174801, accessed 15 August 2019.

M. Weiser, 1993. “Some computer science issues in ubiquitous computing,” Communications of the ACM, volume 36, number 7, pp. 74–84.
https://doi.org/10.1145/159544.159617, accessed 15 August 2019.

M. Weiser, 1991. “The computer for the 21st century,” Scientific American, volume 265, number 3, pp. 94–105.
https://doi.org/10.1038/scientificamerican0991-94, accessed 15 August 2019.

M. Weiser and J.S. Brown, 1997. “The coming age of calm technology,” In: P.J. Denning and R.M. Metcalfe. Beyond calculation: The next fifty years of computing. New York: Copernicus, pp. 75–85.
https://doi.org/10.1007/978-1-4612-0685-9_6, accessed 15 August 2019.

Y. Zhang, L. Song, C. Jiang, N.H. Tran, Z. Dawy, and Z. Han, 2017. “A social-aware framework for efficient information dissemination in wireless ad hoc networks,” IEEE Communications, volume 55, number 1, pp. 174–179.
https://doi.org/10.1109/MCOM.2017.1600029CM, accessed 15 August 2019.

 


Editorial history

Received 19 February 2019; accepted 15 August 2019.


Creative Commons License
This paper is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Ubiquitous tools, connected things and intelligent agents: Disentangling the terminology and revealing underlying theoretical dimensions
by Katrin Etzrodt and Sven Engesser.
First Monday, Volume 24, Number 9 - 2 September 2019
https://journals.uic.edu/ojs/index.php/fm/article/view/9700/8071
doi: http://dx.doi.org/10.5210/fm.v24i9.9700





A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2019. ISSN 1396-0466.