Executive Summary

In this paper, the healthcare revolutionizing technology of Smart Spaces is defined, examined, and implemented across three use case scenarios: Smart Gym, Smart Hospital, and Smart Home. In each use case, various technological aspects (IoT, AI, VR/AR, etc.) are explained to demonstrate a Smart Space solution. Further, each use case explains the core benefits and concerns. Finally, each use case draws to light some gaps in current research and technologies and presents how a Smart Space could be applied to an under-researched area.

Introduction

“Healing is a matter of time but it is sometimes also a matter of opportunity” – Hippocrates, 5th Century BCE

Nearly all doctors recite the Hippocratic oath. However, shouldn’t Hippocrates’ words of wisdom transcend the realm of medicine? If Hippocrates was alive today, would he not have a lot to say about how technology can help transform the practice of medicine? The answer to both of these questions is undoubtedly, yes. If this is the case, then what types of new and emerging technology should be applied to healthcare to increase health, wealth and happiness? The answer is that there is a plethora of cases where information technology and healthcare can and should be merged. Interestingly, there are almost too many use-cases given the emerging tools of Artificial Intelligence, Machine Learning, Internet of Things (IoT), Virtual Reality, Augmented Reality, and Blockchain. All will play a huge role in healthcare over the next decade and beyond. However, this paper will focus on the concept of Smart Spaces within healthcare. Specifically, this paper will examine what a Smart Space is and different use-case scenarios for how it can be applied to enhance healthcare.

Thesis: By combining emerging technologies, Healthcare Smart Spaces will lower costs, improve stakeholder experience, and optimize healthcare outcomes.

 

Method of Analysis

The method used to explore our thesis was to research available vetted materials from reputable publications and sites. Specifically, material was accessed from various journal databases using the Virginia Tech Library system. After researching, the chosen materials were analyzed to discover details that could be used for a discussion on Smart Spaces, Smart Gyms, Smart Hospitals, and Health Smart Homes. Finally, our unique contributions were added to help further the idea of Smart Spaces in Healthcare.

 

Literature Review

Literature review is an integral part of our research. The authors in Table 1 contributed to the formation of our Smart Space Definition:

Table 1: Smart Space Definition Contributing Authors

Author(s)	Research Highlights
Atzori, Luigi, Antonio Iera, and Giacomo Morabito	Internet of Things definition, potentials, and societal roles
Dmitry G. Korzun, Ivan V. Galov, Aleksandr A. Lomov	Smart space service recognition, construction, and perception
Sergey A. Marchenkov, Dmitry G. Korzun, Anton I. Shabaev, Anatoly V. Voronin	Open source middleware as an enabler of smart space implementation in IoT environment
Xiaoyang Wu et al.	Smart space privacy management including visibility management in healthcare setting

 

IoT and Smart Spaces

To truly understand what a Smart Space is, the fundamental concept of Internet of Things (IoT) must be understood.

What is IoT?

IoT is a hot topic across many industries as there is a wide variety of potential applications, particularly in the healthcare industry. However, IoT is not cut and dry, as many researchers, institutions, and agencies have varying definitions. Some researchers^[1] point out that there are three key stages, or generations, of IoT: “tagged objects”, “things interconnected through web technologies”, and “social objects, semantic data representation, and cloud of things.” Whereas, agencies like the International Telecommunication Union (ITU) consider IoT as combination of technologies – Radio-Frequency IDentification (RFID) technologies, sensor technologies, smart technologies, and nanotechnology. Together, these technologies allow for device “tagging” (identification), “feeling” (sensing), “thinking” (intelligence), and “shrinking” (size reduction)^[2]. Regardless of the source of the definition, IoT at its core is a network of interconnected smart devices (smart meaning internet capable with local memory, compute functionality, and possibly artificial intelligence (AI)) that continuously communicate, interact, and share data. Further, these devices can be controlled, monitored, and maintained remotely over the internet. Consequently, IoT has opened new opportunities within the healthcare arena, in particular smart spaces.

Based on current research and practice, a Smart Space can be defined as a computerized system that involves multiple users, multiple technologies to sense, construct, and provision services in an intelligent, seamless, reliable, and secure manner. It can be understood as a versatile and extensible platform built upon generic and domain-specific processes and technologies such as APIs, open source tools, and knowledge libraries. In detail, Smart Space can be understood through four aspects:

Smart Space is a service. The requirement gathering, design, deployment, and operations of a Smart Space is centered around providing high-quality, seamless, constant, and secure services to the end users. Smart Space differentiates itself from other IT-enabled services by emphasizing intelligent service recognition (detection of user needs), construction (automated data processing), and perception (information provision to the user for decision-making)^[3].

Smart Space is a cohesive computerized system involving diverse users, devices, and software. It is an orchestrated effort between users, system administrators, sensors, devices, software platforms, subsystems, and robust data flows to provide the services promised to end users. Existing researchers also refer Smart Spaces as ecosystems, environments, and platforms to convey the same meaning.

Smart Spaces are versatile and interoperable. Smart Spaces have broad and deep use cases in various industries particularly sport/fitness, managed care, and clinics providing health care. Despite the recent progress in Smart Space, there is significant opportunity to create both generic and domain-specific knowledge libraries, APIs, and open source tools to address the common and unique needs in deploying Smart Space. For example Sergey A. Marchenkov suggests using open source middleware to enable the implementation of the smart space concept in IoT environments^[4].

Finally, Smart Spaces are subject to reliability and security challenges that most IT systems face. The constant interaction between users and other participants of Smart Spaces put high pressure on the reliability and system security. Researchers including Xiaoyang Wu has suggested solutions to address such challenges by ensuring privacy management including visibility control^[5].

 

With this background on Smart Spaces, various Use Case Scenarios can now properly be discussed regarding the combination of Healthcare and Smart Spaces.

 

Use Case Scenario 1: Smart Gyms

Very little data exists on combining the ideas of Smart Spaces with Gyms. Presumably, this provides a great opportunity for those wishing to pursue this idea of ‘Smart Gyms’. In fact, it is not difficult to extract what has been imagined with ‘Smart Homes’ and apply it to the gym environment.^[6] Smart Gyms provide an atmosphere where a gym user seamlessly engages with IoT gym equipment that records their workout data and that data is presented it in a meaningful way, i.e., Workout Dashboard. The key here is a constant interaction between the humans and the surrounding environment that is captured by the hardware onsite. This interaction is captured and processed by various software that help reduce costs, improve the gym experience, and increase overall health outcomes.

 

1. Hardware

Previous research, shown in Table 2, has identified a variety of methods to capturing human behavior data within an environment:

Table 2. Previous Research on Smart Space Hardware

Author(s)	Detection Device(s)
Mohan and Vittal[7]	Grid of ultrasonic sensors
Mannini, Intille, Rosenberger, Sabatini, and Haskell[8]	Single accelerometer sensor
Chawla and Wagner[9]	Single chest-mounted accelerometer system
Vaizman, Ellie and Lanckriet[10]	Smart Phone and Pebble Smartwatch
Davis, Owusu, Bastani, Marcenaro and Feijs[11]	Waist-mounted smartphone belt
Ruan et al.[12]	RFID tags embedded within a room
Cristani, Raghavendra, and Bue[13]	Audio/video cameras

 

Smart Gyms will focus mainly on advanced depth cameras to do most of the data acquisition. In addition, an early prototype of Smart Gyms would also include other data inputs such as mobile devices, wearables, and IoT enabled exercise equipment. This technology hardware already exists or is in an emergent stage. The lynchpin is having a successful fusion of them. Each playing a key role that compounds when combined together.

 

Mobile Device: Perhaps the most important hardware concerning end users. Given the widespread adoption of Smartphones in advanced economies, it is safe to presume that every user within a Smart Gym own one. These phones will give users an interface to connect with the other IoT equipment and will provide exercise tracking/analytics.

 

IoT Exercise Equipment: Many exercise equipment, especially cardio related, use some form of a dashboard to track exercise data. The key would be making it transferable to other systems, such as a mobile device or database. However, there is an opportunity to use some sort of unique identifier on the exercise equipment, e.g., QR Code, in order to connect users to them. An interface system would then allow seamless interaction.

 

Depth Sensing Camera: High-end, emerging technology cameras such as Intel’s RealSense Depth Cameras, will help map the smart gym environment. These are not your run-of-the-mill cameras as they use multiple sensors that add depth perception capabilities (See Figure 1).

 

Wearables: Smartwactches in 2019 reached worldwide shipments of 74 million units.^[14] These will play a small role in Smart Gyms initially but as the technology matures and there is increased market growth, it is undeniable that wearables will play a bigger role in Smart Gyms long term.

 

2. Infrastructure Software

Software is what brings together all the hardware and makes a Smart Gym reality. Luckily, a healthy record of literature exists regarding IoT and healthcare, shown in Table 3.

Table 3: Existing Literature regarding IoT and Healthcare

Author(s)	Research Highlights
Bhatia and Sood[15]	Proposed Cloud Centric IoT that assessed health related attributes of a trainee during exercise. Predictive monitoring and visualization of health state vulnerability.
Yang et al[16]	Intelligent Medicine device for home healthcare services
Salem et al[17]	Smartphone based health assessments
Yong et al[18]	IoT based fitness system for monitoring the health status of exercisers.
Ghosh et al.[19]	Implemented a health care system to monitor health conditions of patients in hospitals
Huang and Cheng[20]	RFID technology used for medical nursing system

 

The design of the system would be a web of interconnected pieces. Exercise Equipment Interface, User Equipment Interface, Back End Processing, and Sensors (Figure 2).

 

 Figure 2 – Smart Gym IoT Infrastructure^[21]

User Equipment Interface: As mentioned in the Hardware section above, most users at the Smart Gym will own a smartphone and/or wearable device. These devices will give users a hub to view their individualized data and analytics. This will be both iOS and Android capable. An application will be developed to help users interact with exercise equipment as well. This application should incorporate a QR code interaction model.

 

Exercise Equipment Interface: Almost all exercise equipment in the Smart Gym will utilize some form of interface with the user. A QR code seems like the easiest way to accomplish this. However, other vehicles could be used such as fingerprint scanners or facial recognition software. Regardless of how users login, the equipment will provide some sort of Exercise Dashboard that will present exercise metrics in a meaningful way. This Dashboard will also provide the ability for users to record and watch their form as compared to professional athelete’s form.

 

Depth Sensing Camera + Backend Processing: Emerging technology in cameras can provide a tremendous volume of data input from gym users. As discussed in the hardware section, Intel’s Depth Cameras helps construct a 3D environment. Included are cross-platform open source libraries that can be developed for Smart Gym use cases. In the near future, programs will be coded around the cameras dataset that will be mined by Artificial Intelligence and Machine Learning. Insights will reveal how gym users can improve their exercise form, automatically track weight lifting data such as how much weight was loaded onto the barbell, and suggest how many repetitions to do and how much weight to rack. This information is fed into the User Equipment and Exercise Equipment Interface. Furthermore, these cameras will allow for virtual trainers to coach gym users from anywhere in the world.

 

3. Optimizing Operations

The day to day operations of running a business is a clear area where Smart Gym technology can optimize. Vast amounts of data would be generated passively by the various sensors onsite, e.g., temperature, camera footage, etc. Shelke, Harbour and Aksanli write that management can use mine this data to create more informed decisions about where to effectively allocate staff, what portion of their facility is underutilized, address security concerns and when and where to allocate expensive resources such as power dedicated to HVAC systems.^[22]

 

4. Activity Cognizance Engine
   Recognition of different activities by the computer system provides a vital differentiating factor for Smart Gyms. Using images from the ‘Depth Camera’, the gym’s computer system (Computer), would be able to recognize individual members, know which exercise they are doing, track these exercises, and provide analytics. This is not science fiction as deep neural networks provide very effective image recognition systems.^[23] Convolutional Neural Networks (CNN) is a class of deep neural networks, most commonly applied to analyzing visual imagery.^[24] In addition, it would be more favorable to utilize ‘3D’ CNN due to its ability to extract both time and space features.^[25] 3D CNN, therefore, will provide the dataset model. This data set would then be processed by Google’s TensorFlow, thereby creating the Activity Cognizance Engine that powers Computer.^[26] TensorFlow is used due to it being an open source machine learning library that is ideal for developing models from the acquired Smart Gym data. More research will be needed in this area, especially with an emphasis on IT Security in order to realize the full potential of Smart Gyms.

 

5. Gamification

Gamifying the Smart Gym is a unique differentiator between what the literature has already written about. The data acquired from the sensors and from the user’s devices offers a huge opportunity for mining. One aspect to highly consider is how all this data can be used to motivate gym members to exercise more, thereby reaching their full athletic potential. In fact,

Gamification using wearables has been proven to increase the amount of exercise users will engage in.^[27] The data that is gathered from the IoT devices will be displayed on a live ‘Leader Board’. This Leader Board will be included in the Smartphone application and displayed prominently within the gym (Figure 5 / Figure 6).

   

 

 

 

 

 

 

 

 

Use Case Scenario 2: Smart Hospital

Our research has uncovered tremendous opportunity of applying Smart Space to the hospital setting. Hospital, as B. Spyropoulos^[28] points out, is one of the most complex, intense, and valuable spaces in the modern society. Hospital has been leveraging Information and Communications Technologies (ICT) since the inception of information era. Recently Smart Space promises the combination of limited hospital resources and ubiquitous data and information to make sound decisions in improving healthcare outcomes, improving stakeholders’ satisfaction, and decreasing the total cost of healthcare. We first present the benefits, pros, and cons of a few representative Smart Space technologies in a succinct manner. Next we assess the progress and gaps in the Smart Space research and application in the hospital setting. Finally, we look at a unique, high-value use case scenario: Smart Space application in Laboratory and In-Vitro Diagnostics, an area that is under-researched but of high demand.

2. Technological Frontiers in Hospital Smart Space

Smart Resource Allocation Through Internet of Things (IoT) and Advanced Analytics: Valuable resources such as patient beds^[29], operating rooms, surgeons^[30], nurses, and administrative professionals^[31] are of high demand and limited supply. Bottlenecks widely exist given the interdependency within a highly complex system. By deploying reliable, energy-efficient sensors embedded in devices and patient-wearables, the Smart Space creates a continuous inflow of coordinated patient and environmental data. Advanced analytics through data modeling, cutting-edge algorithms, and computerized decision support functionality unlocks big data’s potential to intelligently assist key decisions such as the routing and inbound logistics to operating rooms and patient transfer across facilities and nearby hospitals. Advanced analytics also provides the researchers with multiple scenarios to reflect real facilities and accommodate unexpected events /exceptions during which Smart Space can assist hospital managers faster and better than traditional systems.

Artificial Intelligence (AI-enabled) Hospital Automation: The definition of Artificial Intelligence (AI) has been going through constant evolution. In a traditional sense, AI can be understood as computerized system’s ability to behave in a manner “unique to” human beings such as learning and reasoning. As AI gradually adopts behaviors traditionally only demonstrated by humans, such as natural language recognition and object classification, this definition is subject to changes. Audio Recognition (AR) and Visual Recognition (VR) are two relatively mature areas in AI. Researchers have deployed AR /VR in multiple use cases. Stanford researcher Albert Haque and colleagues^[32] propose the application of vision-based system in hospitals to track people’s movement within the hospitals including the compliance of hand hygiene. Eva Inaiyah Agustin et al. created a voice recognition system for controlling electrical appliances within the hospital room^[33]. Technology companies such as IBM has already applied such technology in the production environment^[34]. Jefferson Hospital in Philadelphia, PA, United States has enabled smart-speaker voice recognition system to answer patient questions such as scheduling and respond to automatable tasks such as raising the blinds.

Smart Hospital IT Architecture: Researchers and developers have started integrating Smart Spaces and the remaining of the hospital IT infrastructure. Significant progress has been found in the development of hospital IT architecture for example, researchers^[35] utilize a framework of layers of systems and data flows supply medical literature recommended services to patients of mental disorders. Researchers^[36] have identified new network solutions to interface wearable devices with hospital EHR through Narrowband IoT to perform medical processes such as monitoring real-time drop rate and the volume of remaining drug. Researchers^[37] also go beyond the traditional functional boundaries of EHR system to provide predictive analytics through models such as Hidden Markov Model. As researchers^[38] alluded to, Smart Hospital IT architecture evolves around Requirement Analysis Models where useful tools such as UML, GRAI, GIM have been deployed to describe, clarify, document, and deploy desired system requirements.

Smart Patient Tracking and Hospital Security management: Hospitals are responsible for the safety and wellness of the workers, patients, and stakeholders on premise. Unfortunately personnel and expensive assets are often under the risks of security intrusion. Even patients are under risks due to conscious and unconscious decisions. Smart Space is an ideal technological solution that allows the hospital to track personnel, patients, and assets while minimizing the risk of security compromise. Chih-Yen Chen and colleagues^[39] propose  a solution consists of wearable devices and indoor positioning system. The included modules have demonstrated satisfying robustness, low-power-consumption, and high-efficiency tailored to hospital tracking needs. Researchers^[40] also go one step beyond by proposing a system that interconnect every system and object (device) in the building. Various kinds of security management mechanisms are combined to protect Smart Space integrity including Tuple-Based Access Control and security mechanism embedded in Named Data Networking.

Smart Hospital Energy Management: An interesting research topic^[41]  involves the interfacing between smart hospital systems with the building management system (BMS). IoT devices have become a new source of high-energy-consumption, in addition to traditional building energy consumers such as elevators, HVAC, and lighting. Researchers have realized the three-layer of building management – sensor, middleware/automation, and efficient management (decision /aggregation) must incorporate the various components including smart IoT devices. The latter, through the enhancement of Artificial Intelligence and business rules can enhance the efficiency of the remaining of the system. Researchers^[42] encapsulated smart space and green hospital into the concept of “smart and green hospital care”, where energy and material-waste are both addressed by key Smart Space concepts such as Service-oriented-architecture and Mobile IP-networks. 

3. Smart Space Hospital Opportunities and Challenges

Despite significant progress in Smart Space research and practice, the broad and deep implementation of Smart Space requires practitioners to address a few key issues:

Digitization and Ingestion of Patient and Non-patient Data: Smart Spaces require the robust data flows across sensors, data warehouses, and decisional modules. Smart Space designers must put in place secure and efficient mechanisms to capture relevant data points while convincing stakeholders about the needs to convert paper-based records into digital formats and allow the transmission of data within and potentially beyond the walls of the facility. In addition, healthcare leaders including hospital managers and policy makers must carefully weigh the benefits and risks of data sharing and advocating for the broader adoption of IoT security, block chain, and secure access provision to rationalize Smart Space adoption.

Lack of domain specific knowledge: The creation and fine-tuning of Artificial Intelligence is done through Machine Learning process with tremendous amount of domain-specific knowledge. Hospital as a complex system must be appropriately broken down into distinctive disciplines to create meaningful use cases and engage subject matter experts. As an example, IBM has been a leader in applying Watson AI to oncology decision-making. This approach could be applied to tackle other disciplines that require long-term investments in data, governance, and expertise.  

Interoperability across systems: Information islands still exist within a hospital facility, and even within a hospital department. Large scale adoption of Smart Space requires hospitals and healthcare organizations to share resources beyond the boundary of organizations.

Descriptive, Predictive, and Prescriptive analytics: Smart Space has been invaluable in displaying the current state of the hospital through the visualization of patient, instrument, and environmental data. However the long-term success and continual development rely on breakthrough in algorithms that predict future events and prescribe decisions and actions to take before adverse events occur and before valuable opportunities disappear.

Use Case Development: Today’s researchers^[43] clearly point out that Smart Hospitals’ multiple departments share certain infrastructure but also demonstrate unique behaviors. Hospital typically can be organized by the following departments: Emergency and Outpatient, Imaging and Radiotherapy, Surgery Intensive Care Units and Wards, In Vitro Diagnostics and Labs, and other Supporting Units. Through our literature review, currently there is limited to none application of Smart Space technologies to In Vitro Diagnostics Labs. This may be due to high technical complexity, proprietary protocols, and high investment requirements. In the next and final part of Smart Hospital section, we dive into this under-researched topic.

4. Smart Space in Laboratory and In-Vitro Diagnostics (Smart Lab)

According to WHO^[44], Laboratory and In-Vitro Diagnostics (IVDs) are the processes to examine specimens derived from the human body to provide information for screening, diagnostics, or treatment monitoring purposes. Despite its important, integral role within the hospital system, Smart Lab is an under-researched topic in academia.

We have identified a few Smart Lab use cases where Smart Space could add unique values:

Predictive Maintenance and Asset Management: In many modern hospitals, middleware is utilized to seamlessly integrate the diverse IVD instruments with the Laboratory Information System (LIS) to automate sample ordering and processing. Despite such bidirectional communication between LIS and IVD instruments, interfaces don’t typically exist between IVD instruments and an asset management system to track the utilization of instruments, document error codes, identify anomalies through robust algorithms, and provide actionable insights in scheduling maintenance and services. In most cases, IVD instruments are maintained on a routine basis. For example, every six-month, preventative maintenance is performed on a HIV/MTB/HBV molecular testing equipment. There is opportunity in slashing costs and increasing machine uptime by scheduling maintenance according to the sensors’ real time data. Total-cost-of-ownership is reduced, because both labor cost in maintenance and part replacement can be optimized. Furthermore, a smart lab system can automatically trigger a work order based on the anomaly it detects from the operational data of the instruments and therefore minimize the chance of disruption during and even before a break-down occurs. Through Machine Learning, Smart Lab can associate determinants with common machine functional issues intelligently so that a trouble-shooting plan can be provided before the technicians arrive; parts can be ordered early given the preliminary analysis of the Smart Lab. Leveraging biometrics and other security features summarized above, Smart Lab can give access to authorized maintenance personnel to perform the necessary job and allow the lab’s management team  to effortlessly sign-off maintenance work electronically.

Smart Supply Chain Management: One major pain point in today’s hospital labs is the ordering of various types of reagents and consumables (R&C). Smart Lab first ingests the Bill of Materials (BOM) provided by the equipment manufacturers to understand the quantity and ratio of consumables necessary to perform a test. It then asks the lab management the quantity of tests they plan to perform in a certain period (e.g. three-month) and intelligently calculate the number of R&C the lab should order. Smart Lab also corrects the initial entry by comparing planned against the actual testing volumes. For example, a window will pops up and alert the user that the usage pattern shows above than predicted volume and an order needs to be placed sooner than expected. By using the term Supply Chain here, we suggest that Smart Lab extends the scope of R&C management to dynamically monitor warehouse inventory and connect with the manufacturers’ ERP system. For example, a pack of reagent should have end-to-end traceability. Once it is loaded to the instrument, the record /history log should be updated to allow exceptional cases such as quality recall. Such traceability also allows the Smart Lab to notify the manufacturers and warehouses when an inventory replenishment should occur. A fully implemented Smart Lab will send an order to the manufacturer according to consumption, instrument functional status, inventory minimum and maximum stock policy, and other business rules. This minimizes the involvement of lab staff in managing the complex process of inventory management and therefore dedicate more energy to providing care to the patients.

Lab Workflow Advisor: A modern lab should follow a robust workflow procedure, often described as a Standard Operating Procedure (SOP). It also orients staff around a set of Key Performance Indicators (KPIs) to achieve desired outcomes in patient care while minimizing costs. In daily operations, it is not uncommon that operators, due to lack of proficiency, behaves against the set SOP. It is also possible that the workflow doesn’t evolve according to the needs of the testing lab. A Smart Lab would uniquely identify each operator by their ID cards and/or biometrics. It tracks the seniority, hours of operating, education, and other attributes and preferences to provide a customized UI and more importantly, well-informed recommendations before the operators make a mistake. Besides error prevention, a Smart Lab should sufficiently consider the ergonomics of the lab environment and adjust e.g. the bench area, the lighting, and temperature accordingly to the operators’ optimal comfort level. Finally a Smart Lab enables both operators and hospital managers to find areas of improvements through machine learning. Best practices of lab testing will be used to train the AI so that the AI can identify the potential areas of improvement after observing months of real operating data. AI then provide recommendations to reduce total turnaround time, adjust workspace, and “right-size” the number of instruments and testing capacity the lab management really needs. AI doesn’t replace lab management in this sense but complement their daily jobs and routine plans.

Total Patient Diagnostics Experience: Smart Lab is a link in the chain of events to create an optimal patient testing experience. Patient-centric-care requires practitioners to manage the “back-office activities” with the “store-front experience” in mind. Patients require fast turnaround of tests. Patients also demand a secure and robust system to store years’ lab records. Finally, patients are of diverse demographics e.g. education and age. A smart lab not only gives doctors the opportunity to order the right test for the patients; it also intelligently remind the patients to take actions in an outpatient setting. COPE can be linked with Smart Lab to allow the automatic prescription; mobile applications can be delivered to the patients’ cellphone in addition to their prescription information. The mobile application serves as an information portal for the patience but also educational resources and scheduler who sends a friendly reminder to patients when a test is due.

Lab for Public Health: National, state, and municipal governments have taken initiatives in transferring relevant laboratory /hospital test results to government-operated healthcare information system to identify, prevent, and minimize disease outbreaks. For example, New York City has leverage its reporting mechanisms to successfully identify West Nile Virus. Smart Labs seamless integrate private and public sector laboratories by sufficiently addressing security, data, and analytics needs. First, Smart Labs intelligently distinguish patient and non-patient data fields and applying block-chain and other necessary technology to ensure data privacy and integrity. Every instance of testing result is transmitted to the government database according to agreed-upon transmission protocol and format through configuration files such as XML. Minimal human intervention is needed to batch-load records to the public report. Smart Lab allows both the hospital and the government to benefit from the aggregation of testing results. For example, if a disease outbreak is forecasted to happen, appropriate alerts will be pushed to the hospitals with sufficient, actionable insights. For example the Smart Lab can plan resources accordingly including  personnel, testing capacity, and R&C supply chain. Finally, AI-enabled Smart Labs allow robust predictive, prescriptive analytics following the patients’ individual medical history and the government’s health guideline.

 

Use Case Scenario 3: Health Smart Home

As discussed, Smart Spaces is already having a profound impact in healthcare, specifically in gyms and hospitals. One last area we will touch on in this paper has also seen strong applications and use cases for Smart Spaces, Smart Homes. Recently, Smart Homes have become a byproduct of technological innovation, convenience, automation, ease of access, and security. While at first a byproduct, Smart Homes and home automation have turned into its own industry, with major companies producing Smart Homes devices, applications, and services. This is evident, as home builders^[45] are now building houses from the ground up to be ‘Smart’^[46]. But what do Smart Homes and healthcare have to do with each other? This question has been asked, answered, and implemented in a variety of ways already. We will briefly discuss this Smart Home healthcare research, define some of the core technology, and also identify an under-researched and high-demand opportunity. 

1. Key Drivers for Health Smart Homes

Ageing Population and Shortage of Healthcare Professionals: According to the United Nations report there is unprecedented ageing occurring globally due to decreasing fertility and mortality rates. A shocking statistic was presented, emphasizing this ageing concern “In 2018, for the first time in history, persons aged 65 years or over worldwide outnumbered children under age five. Projections indicate that by 2050 there will be more than twice as many persons above 65 as children under five.”^[47] Further, according to the U.S. Department of Health and Human Services, Health Resources and Services Administration (HRSA) there will be a shortage of approximately 24,000 primary care physicians by 2050^[48]. This is reinforced by the Association of American Medical Colleges and their findings on physician shortages as of 2019 which paint an even more worrisome picture, estimating a shortage between 47,000 and 122,000 physicians by 2032^[49].

Eldlerly desire to live at home: A study was done was done by the American Geriatrics Society to measure an elderly patients opinion on living permanently in a nursing home, in which greater than 50% of respondents were extremely opposed, with 26% saying they were “very unwilling” and 30% responding they would “rather die”^[50]. However, as the majority of the elderly prefer to stay and age in the comfort of their own home, this presents a challenge for primary care physicians and family members to monitor the elderly’s health and well-being.

Increasing costs of healthcare/assisted living: Taking care of the elderly has been a family matter for a long time and still is today, with approximately 80% of long term care performed by family members^[51]. This is in large part due to the increasing costs of assisted and long term care for the elderly. Table 4 breaks down the costs of elderly care across a variety of care options.

Table 4: Cost of Care Breakdown 2016 vs 2018

Care Type	Monthly / Annual Cost (2016)[52]	Monthly / Annual Cost (2018)[53]	2-Year $ Increase	2-Year % Change
Nursing Home – Semi-Private Room	$6,844 / $82, 128	$7,441 / $89,292	$597	8.72%
Nursing Home – Private Room	$7,698 / $92,376	$8,365 / $100,380	$667	8.66%
Assisted Living Facility	$3,628 / $43,536	$4,000 / $48,000	$372	10.25%
Adult Day Health Care *	$1,360 / $16,320	$1,560 / $18,720	$200	14.71%
Health Aide **	$3, 280 / $39,360	$3,814 / $45,768	$534	16.28%
Homemaker Services **	$3,200 / $38,400	$3,640 / $43,680	$440	13.75%

* Based on annual rate divided by 12 months and assumes 5 days of care per work week per month

** Based on annual rate divided by 12 months and assumes 40 hour work week

 

As Table 1 highlights, healthcare costs are not insignificant and play an important role in determining care decisions for the elderly, especially in circumstances where health insurance or  medicaid is not an option. This is compounded by the fact that some eldlerly care option costs are annually exceeding the U.S. inflation rate (2.1%), with an average annual increase between 1.5%-3.8%^[54]. According to the Genworth Future Cost Data Tables^[55], by 2033, the national average cost for nursing home – private room will be $156,381.

To summarize, research ^{[56],[57],[58],[59]} indicates four key drivers for Health Smart Home adoption:

• Increase in ageing population and decrease in mortality rates
• Shortage of healthcare professionals
• Elderly preference to live at home versus adult community or care facilities
• Increased care costs for eldlerly

2. Health Smart Home (HSH)

The Health Smart Home (HSH) is the next generation of assisted living for the elderly, the only difference being the elderly can remain at their own home and not be dependent on another human being, but rather home automation. In general, a Smart Home is a dwelling place that contains and/or equipped with a variety of smart devices and smart sensors. A more technical definition, defines a Smart Home as “a special kind of home or residence equipped with sensors and actuators, integrated into the infrastructure of the residence, intended to monitor the context of the inhabitant(s) to improve his or her experience at home”^[60]. The goal of a Smart Home being automating routine activities of daily living (ADLs). This could be turning on/off lights, vacuuming, watching tv, listening to music, starting a vehicle, ordering food, etc. While this might seem as a creature comfort and an excessive/expensive way to get out of household chores, it provides an enormous amount of opportunity for elderly care. Hence, HSH have become increasingly popular among the healthcare industry and well researched. HSH have all the capability and automation of a Smart Home, but with the addition of a context-aware health monitoring system^[61]. In brief, a contextual monitoring system monitors the ADLs of an elderly individual and alerts on deviations from the norm, usually indicating decreasing health or an accident^[62]. This proves extremely valuable and a promising solution to the key drivers as previously discussed.

3. Technology Aspects – IOT and AI

As defined earlier, IoT is a network of interconnected smart devices (smart meaning internet capable with local memory, compute functionality, and possibly artificial intelligence (AI)) that continuously communicate, interact, and share data. IoT is the core of a Smart Space and HSH. But in a context-aware health monitoring system, IoT relies on Machine Learning (ML) and AI to find and identify patterns and make decisions without human interaction. Together IoT and ML/AI enable health monitoring capabilities inside the HSH to create an Ambient Assisted Living (AAL) environment^[63]. An AAL refers to a non-intrusive system that makes independent life easier, most applicable for the elderly, through the use of sensors, devices, and software applications. Specifically, an AAL can be thought of in two parts, IoT and ML/AI. IoT are the smart devices and sensors that gather and share the data, whereas the ML/AI is the software and intelligence piece. The IoT aspect can be broken down further into three core categories of devices and sensors: Personal Sensor Networks (PSNs), Body Sensor Networks (BSNs), and Multimedia Devices (M.D.). Table 5 outlines each of these in more detail. Together, these networks combine to create a Wireless Sensor Network (WSN). A WSN can be thought of as an IoT network, in which all the devices and sensors are interconnected to form a network and often times are controlled via coordinator, which handles data flow, data processing and potential decision making^[64].

Table 5: PSN, BSN, and M.D. Definitions and Examples^[65]

Network Type	Definition	Example Technology
Personal Sensor Networks (PSN)	Environmental sensors, used to monitor the elderly individual, their ADLs, and their living environment. Enabler for context-aware HMS	Pressure sensors, Passive Infrared (PIR), Motion Sensors, Water Flow Sensors, Electrical Current Sensors, Contact Switch Sensors, Fall Detection Sensors, etc.
Body Sensor Networks (BSN)	Wearable devices and sensors to continuously collect health related information (vitals, activity, etc.) on the elderly individual	Biosensors (Smart Watch, Smart Belt, Smart Shirt, etc) Smart Phones, Accelerometer, Gyroscope, GPS, etc.
Multimedia Devices (MD)	Integrated smart devices and sensors that allow for continuous data sharing as an elderly individual interacts with them	Home appliances (Smart TVs, Smart Speakers, Smart Assistants (ex. Amazon Alexa, Google Home), Smart Appliances (Refrigerator, Microwave, Oven), Smart Security (Ring Doorbell, Blink Camera System), etc.

 

As one can imagine, an individual creates a lot of data points in a given day. In this case, if an elderly individual has an HSH, there is a lot of data being collected and analyzed. This is extremely important for the intelligence aspect of an HSH and where the ML/AI aspect really takes over. All this data that is continually being collected and analyzed is used to identify highly repeated patterns (via ML) of an elderly individual. Essentially, overtime this creates a strong baseline of expected, normal, and healthy ADL behavior. If and when an elderly individual deviates from that established baseline (ex. fall down, sleep longer than expected, do not move for longer than normal period of time, decrease or spike in vital signs, etc.), AI, being context-aware, detects change and makes an intelligent decision on how to respond, whether that be to alert a family member, primary care physician, or emergency services. Figure 7 is an architectural overview of an HSH.

Figure 7: Architectural Overview of HSH

All pictures in this diagram are from Google Search and fall under licensed under CC BY 2.0 or self created on Microsoft Visio

 

Contributions of Work and Lessons Learned

 

Contribution to Literature

The individual use case sections call out their contribution to the literature, but the topics that require more discussion are listed here.

 

Smart Gyms: Relatively little has been written in regards to Smart Gyms. Yong provides the closest literature that combines the idea of Smart Spaces and Gyms.^[66] This paper contributes the idea of relying heavily on Depth Sensing Cameras to capture data points. This contrasts with what has already been written in which several of the Smart Space scenarios use quite a variety of sensors. However, using one main data gathering device will reduce costs and increase the processing time for the Activity Cognizance Engine.

 

Reduction in costs will come from using 1 main sensor point (Depth Sensing Camera) instead of the usual collection, e.g., grid of ultrasonic sensors, accelerometer sensors, etc. This follows a similar strategy to Tesla whose cars rely mostly on highly advanced cameras as opposed to cameras + lidar technology. When presenting at “Autonomy Day” Elon Musk denounced the industry trend of having both cameras and lidar and that lidars “expensive sensors that are unnecessary. It’s like having a whole bunch of expensive appendices. One appendix is bad, now we’ll put on a whole bunch of them. That’s ridiculous. You’ll see.”^[67] Similarly, this paper’s Smart Gyms require its Depth Sensing Cameras to do the heavy lifting with data acquisition. Using less data points will also increase processing time and require less development since open source libraries already exist for Depth Sensing Cameras. Still, further research will be needed.

 

Health Smart Homes: As discussed, Health Smart Homes and Health Monitoring Systems have been extensively researched for applications regarding the elderly and enabling a continued, healthy independent lifestyle. However, the application and benefits of HSH can also be applied to patients who need light care and are largely independent post-discharge from the hospital. A prime example, is a patient recovering from surgery or injury who can largely function independently but would benefit from the capabilities and monitoring of a Health Smart Home to aid in recovery and daily tasks. With continued, post-discharge monitoring and automated assistance via smart devices and sensors, patients could be better enabled to recover independently and safely. However, further research is still needed.

 

Conclusion and Future Work

Smart Spaces is the next step in blending cutting-edge technology with current healthcare processes and procedures in order to lower costs, improve stakeholder experience, and optimize healthcare outcomes. More than that, Smart Spaces presents an enormous opportunity to reinvent healthcare at an unparalleled level by decreasing overhead costs, streamlining manual process through automation, and increasing the patient experience with shorter wait times, better care and monitoring, and increased visibility, ownership, and enjoyment over their health. However, more work is required across many avenues of Smart Spaces, such as security and data privacy.

 

Security will be a primary concern regarding Smart Spaces given the sensitive nature of the data collected. This paper touches upon security briefly, but more depth is needed. Al-Rabiahh and Al-Muhtadi write that the dynamism, sensitivity, privacy and availability of context information make smart space[s] and its sensors more vulnerable to attack.^[68] Additional security is demanded in Smart Spaces such as trusting context providers and context receivers, protecting context information from disclosure/manipulation.^[69] The downstream effects of not securing the data is a complete loss of faith with the organization which relies on its data expertise as its source of wealth. Applying the learnings from “Security Solutions for Smart Spaces” would help accomplish the goal of comprehensive security.^[70]

 

 

Appendix

 

References

• Al-Rabiaah, Sumayah, and Jalal Al-Muhtadi. “ConSec: Context-Aware Security Framework for Smart Spaces.” 2012 Sixth International Conference on Innovative Mobile and Internet Services in Ubiquitous Computing, 2012, doi:10.1109/imis.2012.41.
• Bhatia, Munish, and Sandeep K. Sood. “An Intelligent Framework for Workouts in Gymnasium: M-Health Perspective.” Computers & Electrical Engineering, vol. 65, 2018, pp. 292–309., doi:10.1016/j.compeleceng.2017.07.018.
• Chawla, M. Wagner, “Using machine learning techniques for user specific activity recognition”, INC, pp. 25-29, 2016.
• “Cost of Long Term Care by State: 2018 Cost of Care Report.” Genworth. October 9, 2018. Accessed July 07, 2019. https://www.genworth.com/aging-and-you/finances/cost-of-care.html.
• “Costs of Care.” LongTermCare.gov. Accessed July 07, 2019. https://longtermcare.acl.gov/costs-how-to-pay/costs-of-care.html.
• Amendola, S., Lodato, R., Manzari, S., Occhiuzzi, C., & Marrocco, G. (2014). RFID Technology for IoT-Based Personal Healthcare in Smart Spaces. IEEE Internet of Things Journal, 1(2), 144-152. doi:10.1109/jiot.2014.2313981
• Cristani, R. Raghavendra, A. Del Bue, V. Murino, “Human behavior analysis in video surveillance: A social signal processing perspective”, Neurocomputing, vol. 100, pp. 86-97, 2013.
• Davis, E. Owusu, V. Bastani, L. Marcenaro, J. Hu, C. Regazzoni, L. Feijs, “Activity recognition based on inertial sensors for ambient assisted living”, Information Fusion (FUSION) 2016 19th International Conference on, pp. 371-378, 2016.
• Demir, Eren, Erdem Köseoğlu, Radosveta Sokullu, and Burhan Şeker. “Smart Home Assistant for Ambient Assisted Living of Elderly People with Dementia.” Procedia Computer Science 113 (2017): 609-14. doi:10.1016/j.procs.2017.08.302.
• M. Ghosh, D. Halder, S.K.A. Hossain, Remote health monitoring system through iot, in: 2016 5th International Conference on Informatics, Electronics and Vision, ICIEV, 2016
• H. Huang, K.W. Cheng, Rfid technology combined with iot application in medical nursing system, Chem. Eng. News 88 (18) (2014).
• Knight, Drew. “Smart Tech in New Homes: Builders Embrace a Digital Age.” Builders Embrace Smart Tech in New Homes. May 16, 2017. Accessed July 07, 2019. https://www.newhomesource.com/guide/articles/builder-smart-tech.
• Liu, Lili, Eleni Stroulia, Ioanis Nikolaidis, Antonio Miguel-Cruz, and Adriana Rios Rincon. “Smart Homes and Home Health Monitoring Technologies for Older Adults: A Systematic Review.” International Journal of Medical Informatics 91 (July 2016): 44-59. doi:10.1016/j.ijmedinf.2016.04.007.
• Mannini, S.S. Intille, M. Rosenberger, A.M. Sabatini, W. Haskell, “Activity recognition using a single accelerometer placed at the wrist or ankle”, Medicine and science in sports and exercise, vol. 45, no. 11, pp. 2193, 2013.
• Mattimore, Thomas J., Neil S. Wenger, Norman A. Desbiens, Joan M. Teno, Mary Beth Hamel, Honghu Liu, Robert Califf, Alfred F. Connors, Joanne Lynn, and Robert K. Oye. “Surrogate and Physician Understanding of Patients Preferences for Living Permanently in a Nursing Home.” Journal of the American Geriatrics Society 45, no. 7 (July 1, 1997): 818-24. doi:10.1111/j.1532-5415.1997.tb01508.x.
• “Median Cost of Nursing Home, Assisted Living, & Home Care.” Genworth. Accessed July 07, 2019. https://www.genworth.com/aging-and-you/finances/cost-of-care/cost-of-care-trends-and-insights.html.
• Mitesh S. PatelStacey ChangKevin G. Volpp. “Improving Health Care by Gamifying It.” Harvard Business Review, 7 May 2019, hbr.org/2019/05/improving-health-care-by-gamifying-it.
• Mohan, A. Vittal, 2017 9th international conference on communication systems and networks (comsnets).
• Mshali, Haider, Tayeb Lemlouma, Maria Moloney, and Damien Magoni. “A Survey on Health Monitoring Systems for Health Smart Homes.” International Journal of Industrial Ergonomics 66 (July 2018): 26-56. doi:10.1016/j.ergon.2018.02.002.
• “New Findings Confirm Predictions on Physician Shortage.” AAMCNews. April 23, 2019. Accessed July 07, 2019. https://news.aamc.org/press-releases/article/2019-workforce-projections-update/.
• Ng, Andrew. “Convolutional Neural Network.” Unsupervised Feature Learning and Deep Learning Tutorial, Stanford University, deeplearning.stanford.edu/tutorial/supervised/ConvolutionalNeuralNetwork/.
• Normie, Lawrence. “Wireless Sensor Networks for Aging-in-Place: Theory and Practice.” Handbook of Smart Homes, Health Care and Well-Being, January 1, 2016, 457-74. doi:10.1007/978-3-319-01583-5_45.
• Ohnsman, Alan. “Tesla’s Elon Musk Trashes Lidar For Self-Driving Cars, But Waymo Is Rolling Out A New One.” Forbes, Forbes Magazine, 23 Apr. 2019, www.forbes.com/sites/alanohnsman/2019/04/23/teslas-elon-musk-trashes-lidar-for-self-driving-cars-but-waymo-is-rolling-out-a-new-one/#61e521745a9d.
• Potter, Jane F. “Aging in America: Essential Considerations in Shaping Senior Care Policy.” Aging Health 6, no. 3 (June 2010): 289-99. doi:10.2217/ahe.10.25.
• Ruan, Q.Z. Sheng, L. Yao, L. Yang, T. Gu, “Hoi-loc: Towards unobstructive human localization with probabilistic multi-sensor fusion”, Pervasive Computing and Communication Workshops (PerCom Workshops) 2016 IEEE International Conference on, pp. 1-4, 2016.
• Salem O, Liu Y, Mehaoua A, Boutaba R. Online anomaly detection in wireless body area networks for reliable healthcare monitoring. IEEE J Biomed Health Inform 2014;18(5):1541–51
• Shelke, Sagar, et al. “Building an Intelligent and Efficient Smart Space to Detect Human Behavior in Common Areas.” 2018 International Symposium on Networks, Computers and Communications (ISNCC), 2018, doi:10.1109/isncc.2018.8530988.
• Suomalainen, Jani, and Pasi Hyttinen. “Security Solutions for Smart Spaces.” 2011 IEEE/IPSJ International Symposium on Applications and the Internet, 2011, doi:10.1109/saint.2011.57.
• Suryadevara, N.k., S.c. Mukhopadhyay, R. Wang, and R.k. Rayudu. “Forecasting the Behavior of an Elderly Using Wireless Sensors Data in a Smart Home.” Engineering Applications of Artificial Intelligence 26, no. 10 (November 2013): 2641-652. doi:10.1016/j.engappai.2013.08.004.
• Vaizman, K. Ellie, G. Lanckriet, Recognizing detailed human context in-the-wild from smartphones and smartwatches, 2016
• Weise, Elizabeth. “Amazon’s Alexa Will Be Built into All New Homes from Lennar.” USA Today. May 09, 2018. Accessed July 07, 2019. https://www.usatoday.com/story/tech/news/2018/05/09/amazons-alexa-built-into-all-new-homes-lennar/584004002/.
• “World Population Prospects – Population Division.” United Nations. 2019. Accessed July 07, 2019. https://population.un.org/wpp/.
• Yang G, Xie L, Mäntysalo M, Zhou X, Pang Z, Da Xu L, et al. A health-iot platform based on the integration of intelligent packaging, unobtrusive bio-sensor, and intelligent medicine box. IEEE Trans Ind Inf 2014;10(4):2180–91.
• Yong, Binbin; Xu, Zijian; Wang, Xin; Cheng, Libin; Li, Xue; Wu, Xiang; Zhou, Qingguo. “IoT-based intelligent fitness system”. In Journal of Parallel and Distributed Computing. August 2018 118 Part 1:14-21 Part 1 Language: English. DOI: 10.1016/j.jpdc.2017.05.006, Database: ScienceDirect

 

Endnotes

^[1] Atzori, Luigi, Antonio Iera, and Giacomo Morabito. “Understanding the Internet of Things: Definition, Potentials, and Societal Role of a Fast Evolving Paradigm.” Ad Hoc Networks 56 (December 24, 2016): 122-40. doi:10.1016/j.adhoc.2016.12.004

^[2] Steiner, Wilfried, Flavio Bonomi, and Hermann Kopetz. “Towards Synchronous Deterministic Channels for the Internet of Things.” 2014 IEEE World Forum on Internet of Things (WF-IoT), May 27, 2015. doi:10.1109/wf-iot.2014.6803205.

^[3] Dmitry G. Korzun, Ivan V. Galov, Aleksandr A. Lomov Smart space deployment in wireless and mobile settings of the Internet of Things. The 3rd IEEE International Symposium on Wireless Systems within the conference on Intelligent Data Acquisition and Advanced Computing Systems 2016;:1-2.

^[4] “Sergey A. Marchenkov, Dmitry G. Korzun, Anton I. Shabaev, Anatoly V. Voronin On applicability of wireless routers to deployment of smart spaces in Internet of Things environments. The 9th IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications

2017.

^[5] “Xiaoyang Wu, Chun Wu, Yuanchun Shi Multi-Depth-Camera Sensing and Interaction In Smart Space. IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable

Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovations 2018;718-720.

^[6] Amendola, S., Lodato, R., Manzari, S., Occhiuzzi, C., & Marrocco, G. (2014). RFID Technology for IoT-Based Personal Healthcare in Smart Spaces. IEEE Internet of Things Journal, 1(2), 144-152. doi:10.1109/jiot.2014.2313981

^[7] J. Mohan, A. Vittal, 2017 9th international conference on communication systems and networks (comsnets).

^[8] A. Mannini, S.S. Intille, M. Rosenberger, A.M. Sabatini, W. Haskell, “Activity recognition using a single accelerometer placed at the wrist or ankle”, Medicine and science in sports and exercise, vol. 45, no. 11, pp. 2193, 2013.

^[9] J. Chawla, M. Wagner, “Using machine learning techniques for user specific activity recognition”, INC, pp. 25-29, 2016.

^[10] Y. Vaizman, K. Ellie, G. Lanckriet, Recognizing detailed human context in-the-wild from smartphones and smartwatches, 2016

^[11] K. Davis, E. Owusu, V. Bastani, L. Marcenaro, J. Hu, C. Regazzoni, L. Feijs, “Activity recognition based on inertial sensors for ambient assisted living”, Information Fusion (FUSION) 2016 19th International Conference on, pp. 371-378, 2016.

^[12] W. Ruan, Q.Z. Sheng, L. Yao, L. Yang, T. Gu, “Hoi-loc: Towards unobstructive human localization with probabilistic multi-sensor fusion”, Pervasive Computing and Communication Workshops (PerCom Workshops) 2016 IEEE International Conference on, pp. 1-4, 2016.

^[13] M. Cristani, R. Raghavendra, A. Del Bue, V. Murino, “Human behavior analysis in video surveillance: A social signal processing perspective”, Neurocomputing, vol. 100, pp. 86-97, 2013.

^[14] “Gartner Says Worldwide Wearable Device Sales to Grow 26 Percent in 2019 #GartnerTGI.” Gartner. November 29, 2018. Accessed July 07, 2019. https://www.gartner.com/en/newsroom/press-releases/2018-11-29-gartner-says-worldwide-wearable-device-sales-to-grow-.

^[15] Bhatia, Munish, and Sandeep K. Sood. “An Intelligent Framework for Workouts in Gymnasium: M-Health Perspective.” Computers & Electrical Engineering, vol. 65, 2018, pp. 292–309., doi:10.1016/j.compeleceng.2017.07.018.

^[16] Yang G, Xie L, Mäntysalo M, Zhou X, Pang Z, Da Xu L, et al. A health-iot platform based on the integration of intelligent packaging, unobtrusive bio-sensor, and intelligent medicine box. IEEE Trans Ind Inf 2014;10(4):2180–91.

^[17] Salem O, Liu Y, Mehaoua A, Boutaba R. Online anomaly detection in wireless body area networks for reliable healthcare monitoring. IEEE J Biomed Health Inform 2014;18(5):1541–51

^[18] Yong, Binbin; Xu, Zijian; Wang, Xin; Cheng, Libin; Li, Xue; Wu, Xiang; Zhou, Qingguo. “IoT-based intelligent fitness system”. In Journal of Parallel and Distributed Computing. August 2018 118 Part 1:14-21 Part 1 Language: English. DOI: 10.1016/j.jpdc.2017.05.006, Database: ScienceDirect

^[19] A.M. Ghosh, D. Halder, S.K.A. Hossain, Remote health monitoring system through iot, in: 2016 5th International Conference on Informatics, Electronics and Vision, ICIEV, 2016

^[20] C.H. Huang, K.W. Cheng, Rfid technology combined with iot application in medical nursing system, Chem. Eng. News 88 (18) (2014).

^[21] Some images used from product websites: Intel Depth Sensing Camera, Kinect Mapping, Apple Watch and Squat Rack. Other images were created from MS PowerPoint.

^[22] Shelke, Sagar, et al. “Building an Intelligent and Efficient Smart Space to Detect Human Behavior in Common Areas.” 2018 International Symposium on Networks, Computers and Communications (ISNCC), 2018, doi:10.1109/isncc.2018.8530988.

^[23] Yong at 17

^[24] Ng, Andrew. “Convolutional Neural Network.” Unsupervised Feature Learning and Deep Learning Tutorial, Stanford University, deeplearning.stanford.edu/tutorial/supervised/ConvolutionalNeuralNetwork/.

^[25] Id.

^[26] https://www.tensorflow.org/

^[27] Mitesh S. PatelStacey ChangKevin G. Volpp. “Improving Health Care by Gamifying It.” Harvard Business Review, 7 May 2019, hbr.org/2019/05/improving-health-care-by-gamifying-it.

^[28] B. Spyropoulos TOWARD THE DATA-DRIVEN “SMART” AND “GREEN” HOSPITAL-CARE. 1-9.

 

^[29] “Eva Inaiyah Agustin et al Voice recognition system for controlling electrical appliances in smart hospital room. TELKOMNIKA 2019;17(2):965~972.”

^[30] “Abbas Al-Refaie, Toly Chen, Mays Judeh Optimal operating room scheduling for normal

and unexpected events in a smart hospital. Oper Res Int J (2018);18:579–602.”

^[31] https://www.ibm.com/ibm/history/ibm100/us/en/icons/smarterhealthcare/.

 

^[32] Albert Haque et al. Towards Vision-Based Smart Hospitals: A System for

Tracking and Monitoring Hand Hygiene Compliance. arXiv:1708.00163v3 [cs.CV] 24 Apr 2018; 1-13.

^[33] Eva Inaiyah Agustin et al Voice recognition system for controlling electrical appliances in smart hospital room. TELKOMNIKA 2019;17(2):965~972.

^[34]  https://www.ibm.com/blogs/internet-of-things/cognitive-announcement-jefferson-hospital

^[35]Youjun Li, Zhijiang Wan, Jiajin Huang, Jianhui Chen, Zhisheng Huang, Ning Zhong

International WIC Institute, Beijing University of Technology A Smart Hospital Information System for Mental Disorders. IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology 2015;1-4

^[36] Haibin Zhang , Member, IEEE, Jianpeng Li, Bo Wen, Yijie Xun, and Jiajia Liu, Senior Member, IEEE Connecting Intelligent Things in Smart Hospitals Using NB-IoT. IEEE INTERNET OF THINGS JOURNAL 2018;5(3):1-11.

^[37] “Debojyoti Seth et al. Hidden Markov Model and Internet of Things Hybrid driven Smart Hospital. 8th ICCCNT 2017;1-7.

^[38] Najeh Lakhoua, et al. Review on Smart Hospital Management System Technologies. Research and Science Today 2019;1(17):1-19.

^[39] Chih-Yen Chen Implementation of Wearable Devices and Indoor Positioning System for a Smart Hospital Environment. IEEE 2018;1-5.

^[40] Andreas P. Plageras et al. Solutions for Inter-connectivity and Security in a Smart Hospital Building. IEEE 2017;1(17):1-6.

^[41] Andreas P. Plageras et al. Solutions for Inter-connectivity and Security in a Smart Hospital Building. IEEE 2017;1(17):1-6.

^[42] B. Spyropoulos TOWARD THE DATA-DRIVEN “SMART” AND “GREEN” HOSPITAL-CARE.1-9.

^[43] B. Spyropoulos TOWARD THE DATA-DRIVEN “SMART” AND “GREEN” HOSPITAL-CARE. 1-9.

^[44] https://www.who.int/in-vitro-diagnostic/en/

^[45] Weise, Elizabeth. “Amazon’s Alexa Will Be Built into All New Homes from Lennar.” USA Today. May 09, 2018. Accessed July 07, 2019. https://www.usatoday.com/story/tech/news/2018/05/09/amazons-alexa-built-into-all-new-homes-lennar/584004002/.

^[46] Knight, Drew. “Smart Tech in New Homes: Builders Embrace a Digital Age.” Builders Embrace Smart Tech in New Homes. May 16, 2017. Accessed July 07, 2019. https://www.newhomesource.com/guide/articles/builder-smart-tech.

^[47] United Nations, Department of Economic and Social Affairs, Population Division (2019). “World Population Prospects 2019: Highlights” (ST/ESA/SER.A/423).

^[48]U.S. Department of Health and Human Services, Health Resources and Services Administration, National Center for Health Workforce Analysis. 2016. “National and Regional Projections of Supply and Demand for Primary Care Practitioners: 2013-2025”. Rockville, Maryland.

^[49] “New Findings Confirm Predictions on Physician Shortage.” AAMCNews. April 23, 2019. Accessed July 07, 2019. https://news.aamc.org/press-releases/article/2019-workforce-projections-update/.

^[50] Mattimore, Thomas J., Neil S. Wenger, Norman A. Desbiens, Joan M. Teno, Mary Beth Hamel, Honghu Liu, Robert Califf, Alfred F. Connors, Joanne Lynn, and Robert K. Oye. “Surrogate and Physician Understanding of Patients Preferences for Living Permanently in a Nursing Home.” Journal of the American Geriatrics Society 45, no. 7 (July 1, 1997): 818-24. doi:10.1111/j.1532-5415.1997.tb01508.x.

^[51] Potter, Jane F. “Aging in America: Essential Considerations in Shaping Senior Care Policy.” Aging Health 6, no. 3 (June 2010): 289-99. doi:10.2217/ahe.10.25.

^[52] “Costs of Care.” LongTermCare.gov. Accessed July 07, 2019. https://longtermcare.acl.gov/costs-how-to-pay/costs-of-care.html.

^[53] “Cost of Long Term Care by State: 2018 Cost of Care Report.” Genworth. October 9, 2018. Accessed July 07, 2019. https://www.genworth.com/aging-and-you/finances/cost-of-care.html.

^[54] “Median Cost of Nursing Home, Assisted Living, & Home Care.” Genworth. Accessed July 07, 2019. https://www.genworth.com/aging-and-you/finances/cost-of-care/cost-of-care-trends-and-insights.html.

^[55] “Cost of Long Term Care by State: 2018 Cost of Care Report – Future Cost Data Tables.” Genworth. October 9, 2018. Accessed July 07, 2019. https://pro.genworth.com/riiproweb/productinfo/pdf/289701.pdf

^[56] Suryadevara, N.k., S.c. Mukhopadhyay, R. Wang, and R.k. Rayudu. “Forecasting the Behavior of an Elderly Using Wireless Sensors Data in a Smart Home.” Engineering Applications of Artificial Intelligence 26, no. 10 (November 2013): 2641-652. doi:10.1016/j.engappai.2013.08.004

^[57] Liu, Lili, Eleni Stroulia, Ioanis Nikolaidis, Antonio Miguel-Cruz, and Adriana Rios Rincon. “Smart Homes and Home Health Monitoring Technologies for Older Adults: A Systematic Review.” International Journal of Medical Informatics 91 (July 2016): 44-59. doi:10.1016/j.ijmedinf.2016.04.007.

^[58] Mshali, Haider, Tayeb Lemlouma, Maria Moloney, and Damien Magoni. “A Survey on Health Monitoring Systems for Health Smart Homes.” International Journal of Industrial Ergonomics 66 (July 2018): 26-56. doi:10.1016/j.ergon.2018.02.002.

^[59] Potter, Jane F. “Aging in America: Essential Considerations in Shaping Senior Care Policy.” Aging Health 6, no. 3 (June 2010): 289-99. doi:10.2217/ahe.10.25.

^[60]Liu, Lili, Eleni Stroulia, Ioanis Nikolaidis, Antonio Miguel-Cruz, and Adriana Rios Rincon. “Smart Homes and Home Health Monitoring Technologies for Older Adults: A Systematic Review.” International Journal of Medical Informatics 91 (July 2016): 44-59. doi:10.1016/j.ijmedinf.2016.04.007.

^[61] Mshali, Haider, Tayeb Lemlouma, Maria Moloney, and Damien Magoni. “A Survey on Health Monitoring Systems for Health Smart Homes.” International Journal of Industrial Ergonomics 66 (July 2018): 26-56. doi:10.1016/j.ergon.2018.02.002.

^[62] Suryadevara, N.k., S.c. Mukhopadhyay, R. Wang, and R.k. Rayudu. “Forecasting the Behavior of an Elderly Using Wireless Sensors Data in a Smart Home.” Engineering Applications of Artificial Intelligence 26, no. 10 (November 2013): 2641-652. doi:10.1016/j.engappai.2013.08.004

^[63] Demir, Eren, Erdem Köseoğlu, Radosveta Sokullu, and Burhan Şeker. “Smart Home Assistant for Ambient Assisted Living of Elderly People with Dementia.” Procedia Computer Science 113 (2017): 609-14. doi:10.1016/j.procs.2017.08.302.

^[64] Normie, Lawrence. “Wireless Sensor Networks for Aging-in-Place: Theory and Practice.” Handbook of Smart Homes, Health Care and Well-Being, January 1, 2016, 457-74. doi:10.1007/978-3-319-01583-5_45.

^[65] Mshali, Haider, Tayeb Lemlouma, Maria Moloney, and Damien Magoni. “A Survey on Health Monitoring Systems for Health Smart Homes.” International Journal of Industrial Ergonomics 66 (July 2018): 26-56. doi:10.1016/j.ergon.2018.02.002.

^[66] Yong at 185.

^[67] Ohnsman, Alan. “Tesla’s Elon Musk Trashes Lidar For Self-Driving Cars, But Waymo Is Rolling Out A New One.” Forbes, Forbes Magazine, 23 Apr. 2019, www.forbes.com/sites/alanohnsman/2019/04/23/teslas-elon-musk-trashes-lidar-for-self-driving-cars-but-waymo-is-rolling-out-a-new-one/#61e521745a9d.

^[68] Al-Rabiaah, Sumayah, and Jalal Al-Muhtadi. “ConSec: Context-Aware Security Framework for Smart Spaces.” 2012 Sixth International Conference on Innovative Mobile and Internet Services in Ubiquitous Computing, 2012, doi:10.1109/imis.2012.41.

^[69] Id

^[70] Suomalainen, Jani, and Pasi Hyttinen. “Security Solutions for Smart Spaces.” 2011 IEEE/IPSJ International Symposium on Applications and the Internet, 2011, doi:10.1109/saint.2011.57.