Term Paper: The Ethics of the Internet of Things
Often called the Internet of everything, the term Internet of Things(IoT) was coined in 1999 by Kevin Ashton, who dreams of a system where the Internet is connected to every physical object, via ubiquitous sensors (science direct). Specifically, the definition of IoT, as defined by Sholla, Naaz, and Chisti in their article Incorporating Ethics in Internet of Things (IoT) Enabled Connected Smart Healthcare, is a novel approach to networking that connects even ordinary objects to the global Internet by embedding sensing and communication abilities in them. Generally, this concept of IoT is thought of in the context of Smart Appliances, such as home security systems, smart lights, and other electronics such as Google Home.
In a broader sense however, IoT describes any concept in which “computation and data communication [are] embedded in, and distributed through, our entire environment” (Grayson, 2017). This corresponds to the idea that IoT is not based on a single device. It is all about connecting all devices together using services, allowing vendors to ultimately capture data to be used to evaluate and influence our communities (Grayson, 2017). The US Government Accountability Office (GAO) goes further to specify that the Internet of Things (IoT) refers to the technologies and devices that sense information and communicate it to the Internet or other networks and, in some cases, act on that information. In theory, IoT is the idea of making everyday appliances "smart" by connecting them through a network which then allows the devices to communicate with one another and with the user for a more personalized and nuanced experience. At the core of IoT is the evolving relationship between humans and machines (Sankar, 2014).
Recently, these IoT devices are increasingly being used to communicate and process quantities and types of information that have never been captured before and respond automatically to improve industrial processes, public services, and the well-being of individual consumers (GAO, 2017). New research is being done in the area of IoT, leading to the rise of numerous applications, including smart healthcare, smart transportation, industrial automation. Also, many areas in the world are moving forward with the ideas of smart cities, where more applications and services are being made automated. Moreover, building the IoT has become a global trend of governments across the globe. As the government initiatives form the foundations in creating a smart infrastructure, governments initiate projects to improve infrastructures, and to construct new channels that are more advanced and accessible (Gubbi, Buyya, Marusic, and Palaniswami, 2013).
By 2020, it is projected that 50 billion IoT devices will be connected to the Internet worldwide. By 2015, the use of IoT devices had reached on average 3.5 connected devices per person, and is predicted to reach almost 7 per person by 2020. By 2019, the number of mobile broadband access points worldwide are expected to reach 8 billion. In terms of the global economy, it is predicted that there will be over 4.5 million IoT jobs by 2020. IoT is expected to reach $14.4 trillion of value over the next decade, with $97 billion additional revenue in the Medical device industry by 2024. Similarly, IoT is predicted to increase total sales of clothing and accessories by 70 percent over the 2018 year through incorporating computer technology, rising from $3 billion today to $42.5 billion. Moreover, it is projected that there will be a $3.3 trillion market for ‘Smart City’ applications and services by 2025.
Key Technologies of IoT and Categorizing IoT Devices
The key technologies for the IoT include IPv6, which is responsible for things’ identification, sensor technology, which is responsible for dynamic information sensing, communication technology, and network integration technology, which realizes information transmission and intelligent information processing technology. These are referred to as the four key technologies of the IoT (Gubbi, Buyya, Marusic, and Palaniswami, 2013).
With the influx of IoT devices, there arises an issue of identification. Currently, our personal computers each have a unique identifier when connected to the network: the IP. Specifically, computers now use IPv4. This becomes an issue when we consider how many IoT devices connect to the network. Currently, there are hundreds of thousands of IoT devices connected to the network. However, the number of IoT devices are predicted to greatly exceed this. Moreover, when connecting to the Internet, each device also needs its own unique identifier. Thus, the future of IoT lies in IPv6. In contrast to IPv4, IPv6 uses 128-bit addresses. IPv4 only uses 32-bit addresses. With the much larger number of devices that are predicted to need to connect to the Internet, 32-bit addresses no longer provide enough unique IP addresses to give each device a unique identifier. Thus, instead, IoT uses IPv6 which provide many more unique IP addresses (since addresses have more bits and thus there are more unique orderings of those bits).
Sensing technology is another key component of IoT. Sensing technology is used by IoT devices to detect external environmental signals, from things to heat and sound to facial recognition by AI software. This is then used for IoT decision making processes or is the raw data for data transmission.
It is also important for IoT to have a seamless transfer. Specifically, this refers to the idea that interaction and information flow with IoT context should be effortless with low transaction and low information cost (Ebersold and Glass, 2016). This is partially achieved through software and networking protocols as well as the hardware of networks. Moreover, this communication may divide into short distance wireless communication, and wide-area network, according to the transmission type. By making full use of different network communication resources, Network integration technology provides richer network services for users through flexible and efficient networking ways, according to different application environments and local conditions (Gubbi, Buyya, Marusic, and Palaniswami, 2013).
In the Internet of Things, in order to sense an event, it requires the deployment of many types of different sensing devices to monitor different attributes of the event, and then determine whether the incident occurred, through the integration of sensing data. This is achieved through having an intelligent information processing technology. Specifically, this technology transforms the physical sensing data into the logic data that is easily interpretable. Typically, intelligent information processing integrates intelligent computation, data mining, optimized algorithm, machine learning, and etc.
Even though most IoT devices share these four key technologies, IoT devices still cover a broad spectrum. In general IoT devices can be broken into a few types. The first category is the wearable biometric sensors, which include fitbits and other devices that keep track of a users biometrics. Following that is the category of niche gadgets. This category generally refers to IoT devices that fill a very specific role. An example of this would be Amazon’s Dash Button. The Dash Button is a button that specifies a certain Amazon product, and, if that product runs low, the user can press the button which will order a new batch of that product. Another example is smart shower heads, which have the specific role of tracking water use. These niche gadgets can then coordinated using hub devices such as Amazon Echo and Google Home. Lastly, devices such as tablets and smartphones are categorized into a group of IoT devices that are capable of a multitude of functions (Grayson, 2017).
Current Broad Implementations of IoT (Global Differences)
As mentioned above, building networks of IoT devices has become a global trend of governments across the globe. For instance, there is a city called Wuxi in China that has been able to integrate IoT into its technical architecture of the Future Internet, a project going on in China (Kranenburg and Bassi, 2012). This push for smart cities in China is mostly backed by the government, but implemented with a bottom-up approach. This means that even though the project is backed by the Chinese government, each smart city is implemented in its own way. This allows each local city to implement the connectivity of IoT in a way that suits itself best, however, creates issues when establishing a country-wide network for IoT.
In contrast, in the European Union, IoT work is mostly done using the stakeholder approach of public-private partnerships (which facilitates competition, but is still handled by the government for sake of regulation). However, for connection of IoT devices, the EU has opted for a more widespread network of IoT policy. This broad adoption of IoT protocols can potentially be privacy invading. However, for this reason, the EU has also set up risk assessment procedures. (Kranenburg and Bassi, 2012)
In the United States, IoT development is mostly driven by private investors and typically IoT is developed on local levels for uses such as smart energy and smart city uses. As of 2012, there was no US wide policy for IoT (Kranenburg and Bassi, 2012).
Important Concerns of IoT Implementation
Since IoT is fairly new technology, there are many technical and ethical issues that have to be considered before widespread use and implementation of IoT technology. However, despite that, IoT is considered by many to be the future of technology, and provide many important benefits to society.
Throughout this paper, questions I hope to answer are as follows:
1) Whether the benefits of IoT outweigh the consequences. Specifically, should implementing IoT be similar to the way email was implemented (by implementing the technology and figuring out flaws and fixing them as they arose)?
2) What ethical regulations should be established with IoT (since currently, as it is a fairly new technology, there is a lack of rules and regulations over IoT)?
3) Most importantly, how should IoT be implemented (nationally in contrast to locally or through capitalism, and what other things should we keep in mind when implementing IoT)?
The Good (of IoT)
Futuristic goals of IoT devices include many new innovations in technology. Some futuristic ideas (Anderson and Rainie, 2014) include subcutaneous sensors or chips that provide patients’ real-time vital signs to self-trackers and medical providers, remote control apps that allow users’ phones to monitor and adjust household activities—from pre-heating the oven to running a bath to alerting users via apps or texts when too much moisture or heat is building up in various parts of the home (potentially alerting users to a leak or a fire). Similarly, many advertisers of IoT describe smart cities where ubiquitous sensors and GPS readouts allow for vastly smoother flows of traffic; warnings and suggestions to commuters about the best way to get around traffic– perhaps abetted by smart alarm clocks synched to their owners’ eating and commuting habits and their day-to-day calendars. Public applications of IoT include sensored roadways, buildings, bridges, dams and other parts of infrastructure that give regular readings on their state of wear and tear and provide alerts when repairs or upgrades are needed, as well as municipal trash cans that signal when they need to be emptied. In the the areas of industry, IoT is said to vastly improve productivity in manufacturing at every stage, as supply chain logistics are coordinated. More commonplace IoT devices include paper towel dispensers in restrooms that signal when they need to be refilled, alarm clocks that start the coffee maker, and smart appliances working with smart electric grids that run themselves or perform their chores after peak loads subside.
IoT is expected to make life more efficient for all of society through use of automation. Especially, with the advances in AI, smart machines can recognize many things and be used to make decisions based on data gathered from sensors. Moreover, recently, researchers made a machine that can read brain activity (Anderson and Rainie, 2014). The application of such technology with IoT would highly advance many of our everyday devices.
However, one argument against the fantastic possibilities of IoT is that IoT devices will not be as widespread as many think. Specifically, some argue that IoT devices will only end up being niche toys for the wealthy, or that they will only be used in specific environments, such as prisons, hospitals, and the battlefield.
Yet, an important thing to consider is the impact of said “niche toys” on marginalized groups such as the disabled and disadvantaged. For instance, imagine a smart technology that through IoT can tell a blind person about the world around them. Along those same lines, IoT devices can be developed to teach the deaf how to speak that can also differentiate sounds and conversations around the user to provide them a chance to attempt to provide them with a better life. Or, consider the concept of smart machines that can evaluate the slope and how uneven a path is, and react to that based on a database of other people having walked on that path in the past, and use this data to extrapolate how a smart wheelchair should travel across that path or if it should take another route. Moreover, through public implementation of IoT, it is possible to create smart devices that teach and educate the general public. This would then allow typically disadvantaged groups like the poor to learn new things.
A major field of IoT is the idea of “smart” Healthcare. The idea follows along the idea that simple sicknesses that can be solved with a scan and a standardized blood test can be automated, thus leaving issues that need higher training can be prioritized to more highly trained staff. Similarly, monitoring sick people can be more automated which can allow for better identification of when a patient’s statistics become irregular (as the patient is directly connected to medical machinery that tracks their data rather than needing someone to check up every now and then). Also, through the use of IoT, doctors can get patient data faster and IoT devices can optimize which doctors see which patients, such that patients that match doctors’ fields of expertise based on data from a scan (Kranz, 2018).
IoT also could be used to help the environment. Smart applications can manage the user’s use of natural resources to establish user awareness and provide for less wasted resources. For example, smart taps can keep track of how much water a user is using and control the amount of water they waste. Similarly, if one forgets to turn out the lights at home and remember, can just turn them off using smart devices rather than waiting until they return home before turning off lights. Specifically, according to Steven Cristoll, who is a founder and managing partner of Strategic Harmony® Partners and an Advisor in DBI Network, a global management consultancy, it is important to consider the behavior that the product will stimulate downstream after sourcing, manufacturing and transport – in customer/consumer use and at end of life, especially considering a device’s ability to save energy and water or radically alter downstream behavior to save energy or water, reduce waste, positively impact land use or biodiversity, or reduce non-GHG pollution and the toxic load on human health.
The most tangible benefit of IoT is its convenience. The concept of smart applications often include the idea that someone can check their email for work-related issues while doing laundry or making breakfast, which can increase efficiency. Also, according to Kranz, other applications of IoT include networking between cars to prevent or reduce traffic along routes and using location beacons and trackers to more easily locate lost items. Moreover, smart home devices (such as Google Home) can keep track of multiple user devices and can automate many parts of a user's lifestyle (i.e. scheduling lights and when certain devices come on and perform certain actions).
Many companies already use IoT in their manufacturing. Through use of smart devices, manufacturers can see use of energy and keep track of supply-chain statistics and optimize performance and make the process more efficient in terms of time and resources, which improves general society. Specifically, according to the British Computer Society (BCS), “As these embedded devices establish complex networks of human and non-human actors in our public and private spaces, they have the potential to create new relationships between people and computers. Some stated benefits of the IoT include higher business productivity, increased energy and transport efficiency, and greater control and auditing capacity in manufacturing and supply processes.”
Moreover, according to a study in South Korea (Gubbi, Buyya, Marusic, and Palaniswami, 2013), the Internet of Things is becoming a major accelerator for innovation in all industries in South Korea. Specifically, Korea’s IT Industry has been actively working with the government in the area of technologies and equipment. Participating industrial leaders include Samsung, LG, Korea Telecom, and SK Telecom, which all regard IoT as an opportunity to develop new markets, particularly good applications.
Lastly, the IoT provides ways of extending and augmenting human agency (Sholla, Naaz, and Chisti, 2017). Through the interconnectivity of smart devices, a user is able to be more efficient by specializing and automating roles for IoT devices. Similarly, this leaves more time for the user to do other things.
Issues with IoT
One major concern, usually mentioned as the biggest concern of IoT, is security. In a study by the US Government Accountability Office, the GAO found that in 2016 hundreds of thousands of weakly-secured IoT devices were accessed and hacked, disrupting traffic on the Internet. Specifically, according to Sankar, a major concern that IoT is being developed rapidly without appropriate consideration of the profound security challenges involved, and the regulatory changes that might be necessary. Typically, companies rush to push out IoT devices to be trendy and to build brand; however, there are quite a few security holes if the technology isn't properly tested. An issue of IoT being an exciting new fad is that it may cause miniaturization and invisibility of issues, where issues of concern fall under the radar, especially in the push for new and exciting technology (Ebersold and Glass, 2016). This is demonstrated by how IoT products are often sold with old and unpatched embedded OS and software. Also, purchasers often fail to change the default passwords on devices, or fail to set sufficiently strong passwords if they do so.
This security issue is then exacerbated by the typical lack of updates in IoT. An ineffective or nonexistent plan for deploying security updates can be a huge impediment for IoT. Similarly, Sankar warns of things that should to be considered for IoT security: embedded operating system and software security, embedded firewalls, data security when data is at rest and during transit across the network, data encryption, and secure development practices.
The problem of security is especially bad for IoT devices because of the physical impact a compromised IoT device can make. For instance a 2017 US GAO report, remarked on the poor security of IoT devices, citing that in 2015, hackers gained remote access to a car through its connected entertainment system and were able to cut the brakes and disable the transmission. Similarly, hackers can gather data on electrical or water use in a smart home and use that to predict when the user is away for break-ins. Or, consider a smart lock being hacked to let in intruders. According Sankar, In the late 2000s, scientists demonstrated a WiFi pacemaker hack with which one could take control of the device. The threat was so deadly, that former US Vice President Dick Cheney’s doctors disabled his pacemaker’s wireless capabilities to thwart possible assassination attempts. Even government funded IoT networks are not exempt from holes, due in part to the underdeveloped IoT connection protocols (which will be mentioned in detail later). In 2010, Iran nuclear experts raced to stop the spread of the Stuxnet computer worm: a 500-kilobyte computer worm that compromised the software of at least 14 industrial sites in Iran, including a uranium-enrichment plant. In 2014, BBC assembled a team of 7 computer security experts whom demonstrated hacking a home full of smart devices. The computer experts managed to hack every single device and were also able to bug a living room through the microphone on a smart TV.
Finally, the IoT’s lack of good security can be used in Distributed Denial of Service (DDoS) cyber-attacks. Firstly, IoT devices can be hacked to make a botnet (aka thingbot), a group of hacked computers, smart appliances, and Internet-connected devices that have been compromised and can be remote controlled for illicit purposes. Then, these botnets can later be used to perform DDoS attacks. This works by having the malicious controller of the botnet commanding each bot to attempt to access a server. This could make that server unaccessible or crash due to the high amount of traffic.
IoT raises another concern: the question of privacy and trust. In general, IoT devices as discussed above are effective due to large amounts of data mining and processing. This is because IoT devices can sense and save a multitude of data. Moreover, smart devices that monitor public spaces may collect information about individuals without their knowledge or consent. This becomes especially important since smart devices such as fitness trackers collect personally identifiable information like names, email addresses, and dates of birth of users.
Sankar raises a question about trust in the government. He remarks, “chances are Big Data and the Internet of Things will make it harder for us to control our own lives, as we grow increasingly transparent to powerful corporations and government institutions that are becoming more opaque to us.” Gubbi, Buyya, Marusic, and Palaniswami support Sankar’s idea of IoT becoming a means to control society, pointing out that “while ubiquitous and wireless technologies are developed to enable new ways of working, to increase safety and to facilitate coordination, they may interfere with established work practices, undermine productivity and individuals’ satisfaction, and have an unforeseen impact on relations of power and control.” This raises the question of how much of one’s privacy should one relinquish for the sake of new IoT technology, and similarly, do we trust our government enough to trust that IoT and our data will not be abused.
Ebersold and Glass describe this as the ubiquity and pervasiveness of IoT: the user is engulfed and immersed by IoT, with no clear way to opt out of the IoT. Moreover, they discuss the unpredictability and uncertainty of this new technology, warning that new developing technology could lead to emerging behaviors without user having full or even relevant knowledge of the IoT environment.
In the large-scale implementation of IoT, the issue of scalability arises. According to Kranenburg, IoT is powerful due to being a connected mesh of many devices. This puts a burden on our networks to become able to sustain many more interconnected objects (which will outnumber the current number of devices on our networks by several orders of magnitude. On top of that, Kranenburg questions “how this will change our general internet, use of materials and things that will be affected if we start putting silicon chips in everything (maybe use of non-silicon substrates for developing smart components to avoid a dependence on silicon).” Similarly, device communication methods may have to change due to how devices work in IoT connections.
A related issue to scalability of IoT implementation is the rise of the cost of living. This because the development of IoT has become a race between the speed of evolution of IoT and the rate of drop in hardware cost. However, most new IoT technology is still fairly expensive, and since most IoT systems have many parts, for someone to buy an entire IoT system, they would have to spend a lot of money.
Another issue is the lack of standards for IoT protocols. Through their 2017 study of IoT, the GOA set forth the standard that IoT devices and systems must be able to communicate easily. Technical standards to enable this communication will need to be developed and implemented effectively (GOA). However, according to Sankar, this is an issue called the ‘IoT Protocol War.’ In 2013, Qualcomm developed a communication protocol for IoT called AllJoyn. On the other hand, Open Interconnect Consortium (OIC) is an alliance founded in 2014 by Intel, Samsung and Broadcom defining its own connectivity and interoperability standards for IoT. However, a few months later Broadcom left the group over a disagreement on how to handle intellectual property. Google also created a new networking protocol called Thread that aims to create a standard for communication between household devices. Thread does not compete with AllJoyn or OIC as it is a networking protocol, but there are other competing special purpose low-level protocols created by a diverse array of companies. In summary, there is still no standardization of the communication protocol in the IoT world. This could severely hamper interoperability.
Another issues is that currently, the standard of IP for computers and phones on the web is IP4. However, because of the large quantity of IoT devices, to fully support IoT, there are more IP addresses needed (50 billion IoT devices expected to be connected to the Internet worldwide by 2020), such that IPv6 is the standard for IoT devices. Thus, without widespread (global) adoption of IPv6, successful development of IoT will be a lot harder.
Lastly, the IoT can have possible societal effects as well. Ebersold and Glass mention these societal effects as regulatory confusions and possible regulation/social issues. They list the following as social issues that could arise out of IoT.
1) Ambiguity and ontology – identity and system boundaries (artefacts vs human beings due to tagging)
2) Identification – electronic tagging, and who gets to assign, administer, and manage these identities (federal government in contrast to local government, for instance)
3) Social Issues – Embedded intelligence and extended mind – dependency on smart devices may make people cognitively and socially handicapped (Ebersold and Glass, 2016)
The British Computer Society agrees in that “society needs a disconnect sometime from the AI.” Anderson and Rainie also mention that a related strain of argument ties to fears that algorithms cannot necessarily be trusted to make the appropriate decisions. For instance, Aaron Balick, a PhD, psychotherapist, and author of The Psychodynamics of Social Networking, predicted, “Positive things may be tempered by a growing reliance on outsourcing to technologies that make decisions not based on human concerns, but instead on algorithms (however influenced by our own past choices). We may begin to lose sight of our own desires or our own wills, like many of these drivers who we hear about who, because their GPS told them to, end up in the most unlikely places in the face of all sorts of real-world, contrary evidence. What will happen to our own senses of intuition, let alone our capacity to venture into the unknown, learn new things, and our ability to be still and quiet without being in constant relationship to one device or another.”
Considerations for Future Development of IoT
The positives and shortcomings of IoT agree on one crucial point: the quality and ability of IoT to do good while avoiding harm revolves around its implementation and development. Through this, we can summarize some important points to stress in the future development of IoT. Firstly, IoT should (1) be accessible as a public good; (2) be sustainable; (3) provide interoperability; (4) facilitate collaboration; and (5) support experimentation.
In accordance, the BCS-OII Forum Report published in March of 2013 points out some key issues of governance and regulation relating to IoT, including:
1) Rethinking data protection policy and institutional changes to cope with IoT’s scale
2) Accountability and liability: increasing or obscuring accountability for failures, data breaches, costs, and responsibilities for failures.
3) Governing devices that will know a great deal about their users, and actuators that can initiate a series of actions, such as in response to sensor readings.
4) Determining who sets what standards will have major implications for business and industry and national technology-led industrial policies.
5) Alignment of local, national, regional, and global practices and policies
Moreover, the BCS report recommends for IoT implementation to follow the concept of data minimization. Following data minimization, it becomes critical that we implement IoT devices and databases that do not keep data for longer than needed, and don’t track data not needed by the device. Similarly, privacy by design should be a standard of IoT development, to protect users’ privacy and data as much as possible.
Along with careful, ethical development, other needs for the future of IoT include:
1) There needs to be more research on real applications, in order to inform policy and practice. Real-world application trials that involve stakeholders (even on a small scale) are needed to develop more realistic scenarios about potential societal changes, and to distinguish the current reality from hype.
2) There needs to be multi-stakeholder involvement in the early design stages of IoT applications and systems. This allows identification of a wider range of issues that should be taken into account (such as privacy), and is an opportunity to involve the future users of the system.
3) There needs to be greater public understanding and discussion of the technology, its potential benefits, and related issues and challenges.
In summary, there are many important benefits of Internet of Things, from simple devices that make our lives more convenient and personalized to life-saving devices like smart pacemakers. However, at the current state of IoT, the consequences of using and producing more IoT devices currently outweigh the benefits. This is especially since current technology has not developed to the point for IoT to be able to achieve many of its idealized purposes. For instance, the benefit of smart lights in comparison to sensored bridges that can detect its own wear and tear is very small. One provides for greater efficiency and safety in society whereas the other is used for user personalization and convenience.
On the other hand, the dangers of IoT are quite heavy. Privacy and data collection is already a huge issue in our society and IoT would exacerbate that problem by providing more routes through which companies could collect the data of users. Moreover, IoT suffers from weak security, which has very tangible effects if a smart nuclear power plant were hacked, or even a pacemaker which could kill a person.
Specifically, without addressing the issues and shortcomings of IoT, it is not net beneficial for further development and production of Internet of Things devices. Thus, before continuing to develop IoT technology, we should address issues such as scalability and create a more widespread networking and communication protocol to provide for more efficient and secure communication between devices. Similarly, another step would be to develop the applicable government or political policy to address issues of IoT (specifically privacy and consumer consent issues). Without these safeguards in place, IoT could prove to grow out of control.
Typically, with new technology there are two takes on how to go about developing that technology: (1) to just develop it and fix issues as they appear, or (2) to lay down a foundation and set of rules and build off of that. For email, the first method of development was chosen, and many argue that it is important to let IoT grow as it is a new “blank-slate” technology with amazing possibilities. In contrast, I believe it is important to set down rules and a foundation for IoT, since IoT is very different from email in that it could become ubiquitous and IoT sensors could, together, pick up and transmit much more data than email.
Internet of Things cannot just be generalized as a liability, however. It has many possibilities and can provide many new functionalities to society. And, though there are many scary consequences of IoT, we should recall that technology and information security is getting better and better. Similarly, if we follow guidelines in the development of IoT, we can avoid many of these consequences, and perhaps create a guideline in which we should develop and use IoT technology.
...(download the rest of the essay above)