key: cord-198180-pwmr3m4o
authors: Gupta, Deepti; Bhatt, Smriti; Gupta, Maanak; Tosun, Ali Saman
title: Future Smart Connected Communities to Fight COVID-19 Outbreak
date: 2020-07-20
journal: nan
DOI: nan
sha: 
doc_id: 198180
cord_uid: pwmr3m4o

Internet of Things (IoT) has grown rapidly in the last decade and continue to develop in terms of dimension and complexity offering wide range of devices to support diverse set of applications. With ubiquitous Internet, connected sensors and actuators, networking and communication technology, and artificial intelligence (AI), smart cyber-physical systems (CPS) provide services rendering assistance to humans in their daily lives. However, the recent outbreak of COVID-19 (also known as coronavirus) pandemic has exposed and highlighted the limitations of current technological deployments to curtail this disease. IoT and smart connected technologies together with data-driven applications can play a crucial role not only in prevention, continuous monitoring, and mitigation of the disease, but also enable prompt enforcement of guidelines, rules and government orders to contain such future outbreaks. In this paper, we envision an IoT-enabled ecosystem for intelligent monitoring, pro-active prevention and control, and mitigation of COVID-19. We propose different architectures, applications and technology systems for various smart infrastructures including E-health, smart home, smart supply chain management, smart locality, and smart city, to develop future connected communities to manage and mitigate similar outbreaks. Furthermore, we present research challenges together with future directions to enable and develop these smart communities and infrastructures to fight and prepare against such outbreaks.

COVID-19 is an infectious disease caused by a newly discovered coronavirus (SARS-CoV-2) and is rapidly spreading around the world. According to World Health Organization (WHO 1 ), COVID-19 has already affected 215 countries and territories around the world and continue to spread rapidly across other regions. The highly contagious coronavirus outbreak was declared a "pandemic" on March 11, 2020 . WHO 2 reports that the number of positive cases has dramatically surged, with nearly 3 Millions reported cases and 272,000 fatalities as of April 30, 2020 . The fatalities are assuredly the most tragic cost of this disease. In order to control the spread the pandemic, lockdowns, quarantines and stay home orders have been issued by several nations across the globe, which have crippled national and world economy with critical consequences to workers, employers and investors. In addition, the industries, businesses and travel restrictions restrain the supply of goods and services, and the economic disruptions will continue to have a long term impact on global supply chains and economy. In the United States, unemployment rates 3 spiked to 14.7% -its highest level since the Great Depression, in addition to fear of stronger second wave of the disease looming during winter.

Currently, with no cure or vaccine for this disease, the first line of defense to fight against this pandemic is preventative measures and mitigation strategies. As suggested by the WHO, the U.S Centers for Disease Control and Prevention (CDC 4 ) and several other federal organizations suggest personal protective measure (PPE), social distancing, environmental surface cleaning, self isolation, travel restrictions, local and national lockdowns, quarantine, limits on large gatherings, restrictions on opening businesses, and school closures, as some of the preventive measures that are needed to limit the spread of the disease.

However, these guidelines impose restrictions which hinder the way of normal life for humans. It has become a huge challenge to swiftly implement and enforce such measures on a large scale across cities, nations, and around the world. We believe that to effectively enforce and monitor the preventive controls and mitigation strategies for COVID-19, IoT together with its key enabling technologies including cloud computing and artificial intelligence (AI) and data-driven applications can play an important role.

There are several existing examples of use of technology to control the spread of COVID-19 and manage the large gatherings of already infected patients and possibly infected cases. The U.S. CDC has introduced a self-checker 5 application enabled with cloud platform, which helps a patient to make decision to find appropriate healthcare service through questionnaires. However, most people do not have any symptoms who are also known as the silent spreaders/asymptomatic carriers. Therefore, conducting Over the last few years, there has been a huge surge in the number of IoT devices and different types of smart sensors have been introduced. With new technological advancements, this trend is expected to continue and grow in the future. IoT market 13 is currently valued at $267 billion per year and is expected to reach $520 billion by 2021. Another recent article 14 predicted more than 100 billion devices will be Internet-connected by 2025. In today's connected world, not having network capability in a device limits the market potential for that device. As a result, there are large number and various types of network connected IoT devices providing convenience and ease of life to humans. Smart devices have the potential to be a major breakthrough in efforts to control and fight against the current pandemic situation. IoT is an emerging field of research, however, the ubiquitous availability of smart technologies, as well as increased risks of infectious disease spread through the globalization and interconnection of the world necessitates its use for predicting, preventing, and mitigating diseases like COVID-19.

IoT includes large number of novel consumer devices including HDMI sticks, IP cameras, smartwatches, connected light bulbs, smart thermostats, health and fitness trackers, smart locks, connected sprinkler systems, garage connectivity kits, window and door sensors, smart light switch, home security systems, smart ovens, smart baby monitors, and blood pressure monitors. However, mostly these IoT devices are used in a distributed manner based on user and their requirements. IoT technology including smart sensors, actuators, and devices and data driven applications can enable smart connected com-13 https://www.forbes.com/sites/louiscolumbus/2018/08/16/iot-market-predicted-to-double-by-2021-reaching-520b/#82674f91f948 14 https://www.cisco.com/c/dam/en/us/products/collateral/se/internet-of-things/at-a-glance-c45-731471.pdf munities to strengthen the health and economical postures of the nations to fight against the current COVID-19 situation and other future pandemics efficiently.

In this paper, our main goal is to present a holistic vision of IoT-enabled smart communities utilizing various IoT devices, applications, and relevant technologies (e.g., AI, Machine Learning (ML), etc.). Here, we propose a vision of smart connected ecosystem, as shown in Figure 1 , with real-world scenarios in various applications domains with a focus on detecting, preventing and mitigating COVID-19 outbreak.

The major contributions of this paper are outlined below.

• We outline some of COVID-19 symptoms, preventive measures, mitigation strategies, and current problems, challenges and present an overview of adaptable multi-layered IoT architecture and depict interactions between layers focusing on smart communities for COVID-19 requirements.

• We design a smart connected ecosystem by developing multiple conceptual IoT application frameworks including E-Health, Smart Home, Smart Supply Chain Management, Smart Locality, and Smart City. We introduce use-cases and applications scenarios for early COVID-19 detection, prevention, and mitigation.

• We identify and highlight challenges to implement this smart ecosystem including security and privacy, performance efficiency, interoperability and IoT federation, implementation challenges, policy and guidelines, machine learning and big data analytic. Finally, we discuss the interdisciplinary research directions to enable and empower future smart connected communities.

The remainder of this paper is organized as follows. Section 2 discusses the essential characteristics to diagnose, prevent and mitigate COVID-19 disease. Section 3 presents the multi-layered architecture for IoT, whereas Section 4 discusses smart connected ecosystem scenarios in various IoT application domains. Section 5 highlights open research challenges and future directions. Finally, Section 6 draws conclusion to this research paper.

2 Essential Characteristics to diagnose, prevent and mitigate

Coronavirus is transmitted mainly by the infected person's saliva and nasal drips which spread during coughing and sneezing and infects anybody in close contact. Another source of infection is contaminated surfaces in surrounding and high risk areas, such as door handles, railings, elevators, and public restrooms.

COVID-19 is a highly contagious virus with the incubation period stretched from 2 days to 2 weeks after exposure. Symptoms of COVID-19 range from mild symptoms including fever, coughing and shortness of Coping with anxiety disorder, depression issues, and mental health problems.

Shortness of breath or cannot breathe deeply enough to fill your lungs with air, chills.

Avoid face-to-face meetings, practice social distance from other people outside of the home.

Monitor symptoms regularly, wear a cloth covering or N-95 mask over nose and mouth.

Knowledge gaps to understand virus transmission, no specific antiviral treatment, and no vaccine available. Loss of the sense of smell is most likely to occur by the third day of infection and some patients also have experience a loss of the sense of taste.

Cover mouth and nose with a cloth or wear mask when around others, wear gloves and discard them properly.

Manufacturers use of all cleaning and disinfection products, follow the workplace protocol 22 and provide PPE to their employee.

Lack of testing and essential resources such as ventilators, masks, beds, and health staffs, cancel elective surgery.

Diarrhea and nausea a few days prior to fever, CDC says a sudden confusion or an inability to wake up and be alert may be a serious sign.

Cover cough or sneeze with a tissue, then throw the tissue in the trash.

Hospital task force such as increase the number of testing, available the PPE for their staff members, and increase the incentive care. Food and Drug Administration (FDA 24 ) provided emergency use authorization for hydroxychloroquine medicine to treat the people who are suffering from this virus in hospitals. Later, on April 24, 2020 FDA 25 warned against use of hydroxychloroquine to treat this disease outside of the hospital setting or a clinical trial due to risk of heart rhythm problems. People can protect themselves by following some protective measures and help to slow the spread using mitigation strategies. Table 1 provides a comprehensive overview of the symptoms, preventive measures, mitigation strategies and some challenges fighting COVID-19 disease.

One of the easiest preventive measure is to wash your hands frequently and thoroughly with soap and water for at least 20 seconds or use hand sanitizer or an alcohol-based hand rub when soap and water are not available. People should keep social distancing (six feet distance) from others especially from people who are coughing or sneezing. It is suggested to wear mask and gloves in outdoor locations, and to avoid touching the face and surfaces such as the button at a traffic light, a keypad to add a tip for the restaurant take-out order, elevator buttons, etc. Many surfaces are touched by hands accidentally and virus can be potentially picked up and then transmitted to other surfaces and locations. Once the hands are contaminated, the virus can be transferred through eyes, nose or mouth, thus, it enters human body. Respiratory hygiene is another protective measure. There is a need to avoid cough or sneeze into the hands, and to cover mouth and nose with a tissue or bent elbow during cough or sneeze and throw away the tissue immediately. Groceries and packets can be contaminated from coronavirus, and it is recommended to wash grocery items carefully and wipe packets using disinfectant spray. Local public health administrations regularly issue health guidelines, which people should follow. The WHO, governments, and healthcare workers are all urging people to stay home if they can. On top of basic illness prevention, experts said that the best (and only real) defense against disease is a strong immune system. In addition to the physical health, taking care of mental health is also necessary. High stress levels can take a toll on human's immune system, which is the opposite of what people want in this situation.

In addition, mitigation strategies are a set of actions applied by the people and communities (hospitals, grocery stores, and cities) to help slow the spread of respiratory virus infections. These mechanisms can be scaled up or down depending on the evolving local situation. At individual level, if a person is infected with coronavirus, then he/she should self-isolate and follow the guidelines of quarantine provided by hospital. The hospitals must support healthcare workforce, increase testing and intensive care capacity, and availability of personal protective equipment. City governments can appoint task 24 https://www.fda.gov/media/136537/download 25 https://www.fda.gov/drugs/drug-safety-and-availability/fda-cautions-against-use-hydroxychloroquine-or-chloroquine-COVID-19-outside-hospital-setting-or force, open shelters for homeless people, and maintain availability of resources to implement preventive and mitigation strategies for this disease. While each community is unique, appropriate mitigation strategies vary based on the level of community transmission, characteristics of the community and their populations, and the local capacity to implement strategies. Nonetheless, it is crucial to understand the characteristics of this novel virus and spread awareness and up-to-date information across communities through appropriate technology. Consequently, it is essential to address the challenges with significant research and implementation of strategies as shown in the Table. 3 Multi-Layered Architecture

In this section, we explain an integrated multi-layer IoT architecture which can fundamentally change the infrastructure and underlying technologies for smart communities including hospitals, grocery retail stores, transportation, and city etc. as shown in Figure 2 . Our proposed architecture extends and adapts existing IoT and CPS architectures [1] [2] [3] [4] [5] [6] , and focuses on the need of swift enforcement of policies, laws and public guidelines, in order to curtail the widespread of such disease. The architecture integrates a hybrid cloud and edge computing nodes together with IoT and smart sensor devices, to enable real-time and data-driven services and applications needed in COVID-19 pandemics. Overall, the architecture consists of six layers: Object layer, Edge layer, Virtual Object layer, Cloud layer, Network Communication, and Application layer. The object layer is a rich set of IoT devices including sensors, actuators, embedded devices, road side infrastructures, vehicles, etc. These physical objects are spread across and implemented in smart communities like hospitals, retail-stores, homes, parking lots. The edge layer provides local real-time computation and analysis needed for smart resource constrained physical objects. This layer incorporates edge gateways and cloudlets [7] which can enable local computation at this layer overcoming limited bandwidth and latency requirements, and also impacts the usability of the IoT applications.

This multi layer architecture has integrated the concept of virtual objects (VOs) [8] , which are the digital delineation of physical IoT devices. VOs show the current state of corresponding physical objects in the digital space when they are connected, and can also store a future state for these devices when they are offline. Cloud layer provides various services like remote storage, computation, big data analysis, and data-driven AI applications etc. for huge amount of information generated by billions of IoT devices connected to the cloud. We defined a computation layer which comprises of edge layer, virtual object layer, and cloud layer. Computation, data analytic and processing services are performed in this layer.

Network communication layer run among different layers to establish the interaction. It is responsible for connecting physical sensors, smart devices, edge compute nodes or cloudlets, and cloud services with different technologies, and is also used for transmitting and processing sensor data. Application layer delivers specific services to end users through different IoT applications. In the multi-layered smart communities architecture, this application integrates mobile phones, edge computing, cloud computing, AI based analytic, and data-driven services.

This architecture can incorporate the IoT application frameworks within different domains as discussed in the Section 4, and different use case scenarios can be mapped and implemented using relevant technologies associated with each layer of the architecture.

In this section, we discuss various IoT use case scenarios to monitor, diagnose, detect, and mitigate 

It is expected that the global Internet of Medical Things (IoMT) market 26 will grow to a 136.8 billions in

year 2022. As of 2020, there are 3.7 million medical devices in use that are connected and monitor body parameters of the users to inform data-driven applications in making real-time healthcare decisions.

Improving the efficiency and quality of healthcare services in hospitals have been an important and critical challenge during the COVID-19 pandemic time.

In E-health use cases, we will discuss three important scenarios which can help reduce the risk and spread of coronavirus infection. E-health set up comprises of a smart hospital, a remote patient monitoring, and a smart testing booth as shown in Figure 3 . These use-cases involve connected smart sensors, connected devices, robots, patients, hospital practitioners, workers etc. together with IoT applications, edge devices and cloud services to offer data-driven services. Other scenarios including smart pharmacy, smart ambulance and smart parking in hospital's parking area, are also briefly discussed. Such scenarios can be extended with the current proposed architectures described in Section 3.

Various components of smart hospital concept has been studied in the literature [9] [10] [11] [12] [13] . However, it is still a challenge to track COVID-19 patient's record and keep track of essential resources in hospital during such pandemic crisis. The hospitals are overflowing with patients, and running out of hospital beds, PPE and other essential resources needed for treatment and prevention. To overcome these problems, we propose a smart hospital use case here, whihc extends the existing infrastructure to enable coordinated actions for coronavirus patients. Within a smart hospital, RFID sensors can be used to track inventory items like masks, face shields, gauzes, disposable patient examination papers, boxes of gloves, and plastic bottles and vials. These RFID tags also could be an ideal way to keep track of large equipment as well, such as smart beds, ventilators within a smart hospital. Towel, sheets, and blankets must be washed and disinfected regularly, and such items also can be tracked through RFID laundry tags. In Wuhan, Wuchang field hospital 27 provided wearable smart bracelets and rings embedded with multiple sensors to each patient in the hospital, where these IoT devices are synced with the cloud AI platform so that patient's vital signs, body temperature, heart rate and blood oxygen levels, can be monitored regularly by hospital practitioners. In addition, all hospital workers and staff members also wear these smart bracelets and rings to notice any early symptoms of coronavirus infection.

IoT devices generate tremendous amount of data and this data can be collected using edge servers deployed in the hospital facilities. These data sets can be used for training with Federated Deep Learning [14] technique to enhance the intelligence of data-assisted applications which can be used to predict coronavirus infections for hospital practitioners, staff members and also provide insights on coronavirus characteristics and infection trends for the future. Hospital practitioners can check the patient's data through remote applications which will also help in reducing the number of visits to the patient's room to measure her/his vital parameters. In this way, hospital practitioners can not only collect more data in less time with minimal in-person contact but can also reduce the risk for cross-infection from the patients. This technique can significantly help reduce the workload and increase the efficiency of hospital practitioners. Moreover, smart hospitals can utilize smart beds, which sense the presence of a COVID-19 patient and automatically adjust the bed to a good angle if the patient is short of breathe to provide proper support without the need for a nurse to intervene. A Singapore based medical device company 28

invented a smart ventilator, which allows inpatient monitoring process through remote access via an online portal. These ventilators measure the amount of oxygen automatically, or monitor the rate of delivery to the patient, as high pressure to force in more oxygen can damages the lungs of the patient also. These smart ventilators can communicate through patient's smart bracelets with embedded sensors and can respond according to patient's body parameters.

Besides, there are other IoT devices, such as disinfected robots which can autonomously disinfect a patient room regularly and after the patient is discharged, or a specific hospital area post contamination as needed. A study 29 if a patient is in the car. The pharmacist communicates with patient through an application and provides prescribed medicine to them. Before attending the other person, the pharmacist must take some time to sanitize the smart pickup box. In the above scenarios, data and information collected from smart devices are sent to edge gateways, services, or cloud. Due to high security and privacy concerns in health domain, it is important to understand that these edge gateways and cloud-IoT platforms will be owned only by authorized entities, such as hospitals or other highly trusted entities through some private cloud.

We elaborate these challenges in detail in Section 5.

Various aspects of smart home for health have been investigated in the literature [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] . It is nearly impossible to keep everything in the real-world virus-free. However, individuals must exercise due pre- 

The novel coronavirus has people boarded up inside their homes due to lockdowns and stay home orders.

However, people can still go outside for essential services. It is critical to take precautions during outside time and must be aware of possibly infected items and exposure to the virus that they may accidentally bring home. It is paramount to keep our smart homes disinfected and sanitized using different IoT 

During the pandemic, there is a requirement to improvise supply chain management to adapt automatic business processes and also improve the inventory with delivery of essential items. In this subsection, we will discuss two scenarios: smart inventory, and smart retail stores, which illustrate how IoT devices and technologies can enable efficient supply-chain and help in slowing the COVID-19 spread using various prevention and detection mechanisms. The complete scenario of smart supply chain management is shown in Figure 5 .

Various aspects of smart inventory systems have been investigated in the literature [31] [32] [33] [34] [35] [36] [37] . Most of them involve IoT devices where staff use handheld readers to scan the bar code of goods, and then write the storage information to the RFID tags to complete the inventory.

Smart inventory systems show how IoT technology can be leveraged globally to plan and respond under the current pandemic situation. Inventories are facing an unprecedented challenge in coping with the fallout from COVID-19. However, a smart inventory system can provide a safe and secure environment to the workers using IoT technologies. Within the inventories, drones can be used to track all the employees to check their temperature using thermal sensors, and also measure their social distancing practices. Inventory manager can also provide smart wearable devices connected to the centralized cloud to each employee to monitor and track them. If an employee or his/her family member is infected with coronavirus, inventory manager can get notification through data-driven IoT applications. In addition, disinfectant spray can be attached to the shelves and can start to spray when associated sensor senses the sound of sneezing.

California had the earliest stay-at-home order issued on March 19 33 . In California, there has been an early rise in truck activity since the week of March 1. Autonomous delivery robots can also help in smart inventory and help to reduce cross infection. However, truck activity in California has fallen 8.3% from early February. IoT sensors like thermal sensors, GPS, motion sensors can be attached with delivery trucks to maintain the temperature for perishable items and to track the location. This data can be stored on inventory cloud, and can help predict demand and supply for next month. These tools will become the foundation on which supply chain managers gain insight into their markets and erratic supply and demand trends. The RFID antenna scans the number of units on the sales floor and alerts a store manager in case it's low. IoT allows store managers to automate product orders, is capable of notifying when a certain product needs to be re-ordered, gathers data regarding the popularity of a certain item, and prevents employee theft. The retail industry is seeing a rapid transformation, with

The IoT solutions taking the center stage in the sector.

Smart grocery store have been widely investigated in the literature [35, [38] [39] [40] [41] [42] . IoT along with AI and ML technologies can help slow the spread of infection by enforcing prevention and detection mechanisms through connected sensor devices in a smart grocery store. Due to the stay home order, people are panicking and stocking up grocery items. They need to stand in queues for hours outside the store to buy groceries. By employing IoT sensors around the store and wearable IoT devices, a store manager can get a better understanding to slow the spread in the store. Chinese tech firm Kuang-Chi Technologies 34 has developed a smart helmet that is used to identify and target those people who are at high-risk for virus transmission in the retail store. The customer will wear a RFID tag based smart bracelet at the store. At the entrance each customer's smart bracelet will be scanned by a scanner, which will show 33 https://talkbusiness.net/2020/04/groups-share-data-quantifying-COVID-19-impacts-on-trucking-industry/ 34 https://www.yicaiglobal.com/news/chinese-tech-firm-debuts-five-meter-fever-finding-smart-helmet his/her body parameters. If the records shows any symptoms of COVID-19, an alert can be sent to the store manager and individual can be restricted to enter in the store. Similarly, there can be thermal cameras 35 and microphone sensors installed at the store which can detect the people who are coughing in store during shopping and take pictures. These areas will be disinfected and identified individuals may be reported for testing based on their other symptoms and will be categorized as risky customers. This information can be maintained in store for short-period of time to assist in identifying these individuals during their future visits to the store.

From a customer's perspective, the user can enable alerts on his smartphone regarding his grocery list, can see the map of the store and crowded aisles and plan accordingly to maintain social distance while shopping. The customer can visit desired aisles and get items from the smart shelves, which will allow to pick some limited number of items as per family size, and then put items in the smart cart.

Smart shelves will have three common elements -an RFID tag, an RFID reader, and an antenna. Data collected by smart shelves during the day will be analyzed and shopper buying trends, patterns, shopper traffic, etc. will be shared with a store manager to provide customer-related insights to efficiently manage the store inventory and restocking goods.

Most retail stores now allow only ten people at 30 minutes shopping interval slot to avoid large gatherings inside the store. Social distancing can be measured and enforced by autonomous retail worker (robot) together with smart cameras, and smart microphone sensors and speakers to alert the customers, using AI technologies. If two customers come in same aisle and violate the physical distancing norm, an autonomous retail worker will go there to warn them or a loudspeaker attached with smart camera will announce to keep maintain social distancing. Autonomous retail worker can roam around the store and can take note of misplaced items, or products running out of stock (smart shelves also keep track of items and can send alert for restocking as needed). AT&T 36 with Xenex and Brain Corporation has already launched IoT robots to help grocery stores in keeping them clean, killing germs and maintaining well stocked shelves more efficiently. The UVD Robot also uses ultraviolet light to zap infection viruses and sanitize surfaces.

In smart pickup, sensors and other AI based techniques are used to determine whether order is ready to pickup and a person is here to pickup his order. For instance, a parking space or driveway at the store might allow COVID-19 patient or elderly people firstly to avoid waiting time. Smart pickup can automatically allow most vulnerable and infected people first and enforce these rules to inform the vehicles. A restaurant takeout service can follow the same protocol for smart pickup. Intelligent

Transportation System (ITS) [43] can support to deliver resources to essential services and delivery 35 https://spectrum.ieee.org/news-from-around-ieee/the-institute/ieee-member-news/thermal-cameras-are-beingoutfitted-to-detect-fever-and-conduct-contact-tracing-for-covid19 36 https://www.fiercewireless.com/iot/at-t-4g-lte-connects-iot-robots-to-kill-germs-keep-shelves-stocked robot can enhance contact-less delivery, which reduce the spread of the virus. Gupta et al [44] have also elaborated how ITS and smart city infrastructures can be used to enable and enforce social distancing community measures in COVID-19 outbreak.

Smart localities have been widely investigated in the literature [45] [46] [47] [48] [49] [50] [51] . Such localities are a collection of various interdependent human and physical systems, where IoT represents the sensing and actuating infrastructure to estimate the state of human and physical systems and assist in adapting/changing these systems. Here, we discuss two scenarios, which can help humans avoid coronavirus infections and adapt to the 'new normal' in COVID-19 situation living in the smart locality. These two scenarios including other relevant scenarios are shown in Figure 6 .

Every individual who lives in a smart neighborhood will receive notifications regarding allotted time for outside activities, such as riding a bike, a walk on the trail, etc. in order to maintain the social distancing while being outside in the locality common areas. In a smart locality, motion sensors and cameras will sense and count the number of people and send the data to the locality cloud-1 as shown in Figure 6 .

Cloud Smart Analytics 37 service can analyze the locality data and send notification to people regarding 37 https://cloud.google.com/solutions/smart-analytics/ number of positives cases and categorize risk zones with different colors (e.g., red -high risk, yellowmedium risk, and green -low risk) in smart neighborhood. When a person will go for a walk in a smart neighborhood, he/she will receive an alert if any infected person/pet are around and also alert them to avoid high risk (red) zones in smart neighborhood. There will be disinfected sprinkler installed that can spray on the possibly infected areas, such as pedestrian path, common areas, etc. when sensor will sense the presence of infected person in the area through notifications from the locality cloud-1.

The U.S. CDC 38 we have discussed in previous scenarios. Figure 6 shows that the body parameters of the elderly people can be taken through attached body bracelet and sent to edge devices and gateways. IoT devices and applications connected through locality cloud-2 can share information of COVID-19 patients to smart hospitals. The nursing home staff can monitor the patient's body parameters regularly, and will also track other elderly people in the smart nursing home.

In a smart gym, multiple sensors, devices, autonomous devices can be connected through gateway, and gym manager/coach can access the information of each member at different access levels and can communicates to other locality through locality cloud-3. A gym member will receive a notification regarding to come to the gym, it must required to maintain the 25% occupancy at the gym and time interval to sanitize all gym equipment and surface. To enable multi-cloud secure data and information sharing and communications between locality clouds, there is need to be decentralized trust framework in place using advanced technologies like blockchain and trusted distributed computing.

Smart and connected city infrastructures have been investigated in the literature [46, [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] . Daegu 41 has setup a novel system using large amount of data gathered from various sensors and devices, such as surveillance camera footage and credit card transactions of confirmed coronavirus patients to recreate their movements. The Newcastle University Urban Observatory 42 developed a way of tracking of pedestrian, car parks, traffic movement to understand how social distancing is being followed in Tyne and

Wear. However, other major cities need to prepare themselves for coronavirus future outbreak waves. 

Countries have used cell phone data to track citizens' movements during the pandemic 44 showing the geographic data on hot spots and risk zones where people are more likely to get infection.

Smart and connected vehicles have been extensively investigated in the literature [64] [65] [66] [67] [68] [69] [70] [71] . In order to keep patients and healthcare providers safe, drive-thru coronavirus testing sites have been popping up in the city. An autonomous testing vehicle can be used for COVID-19 testing in urban and rural areas. The smart testing vehicle can include infrared body temperature, oxygen level sensors, smart test kit, camera, microphone, and local edge services. It can help reduce exposure of old-age people with preexisting conditions. In these vehicles, a person can enter from one side of the glass-walled area in the car. In-build sensors can record person's body temperature and oxygen level and can store on City cloud-3. It can also provide a test kit to individuals who can test themselves and return it through the car window. In rural areas, autonomous testing vehicles are largely applicable and can help in testing people, as well as inform and make people aware of the COVID-19 pandemic, symptoms and preventive measures.

To flatten the curve of confirmed cases, smart city can provide mass quarantine for coronavirus patients, who have mild symptoms but with higher risk of cross-infection to others. A smart hall or large stadium or facilities can be setup for quarantine with installed sensors, smart devices, robots, and connected to city cloud-2 (as shown in Figure 7 ). Disinfectant robot is an autonomous robot that can sterilize floors in these large areas as discussed in other scenarios. The large-scale disinfectant robot can also be used to clean the roads of the city. Autonomous and self driving vehicles can be used for delivering the post, which will also help reducing the human contact and cutting down the number of COVID-19 cases. Smart city will also provide immunity-based RFID tags to those people, who recover from the disease and allow the tag holders to return to work with extra-precautions. In the future, once COVID-19 vaccines are available, the individual with vaccination can get similar immunity-based RFID tags to prove their immunity.

The development of the proposed smart connected ecosystem requires to address several challenges and needs inter-disciplinary research from an integrated perspective involving different domains and stakeholders. In this section, we will discuss these challenges in detail with examples from each of the proposed scenarios, as illustrated in Figure 8 .

One of the major challenges in the deployment of the smart infrastructure is the security and privacy concerns pertaining to IoT and CPS users, smart devices, data, and applications in different application domains like healthcare, smart home, supply chain management, transportation, and smart city. In health care industry, it is still a challenge to secure connected medical devices and ensure user privacy.

In E-health scenario, for instance, a user visits smart testing booth for COVID-19 testing, and his/her data is transmitted and stored on smart hospital private cloud. Hospitals then share this data with state healthcare staff or city government for tracking and monitoring the user activities. To secure the identity of user and ensure privacy, differential privacy [72] and data masking techniques [73] , such as pseudonymize [74] and anonymize [75] , can be used. However, there are limitations intrinsic to these solutions. In pseudonymize technique, data can be traced back into its original state with high risk of compromising user privacy, whereas it becomes impossible to return data into its original state in anonymize. It is critical to ensure user privacy while deploying IoT and data-driven applications for their wide-adoption in preventing, monitoring, and mitigating COVID-19.

Secure authentication mechanisms including access control and communication control models are necessary for cloud-enabled IoT platforms to defend against unauthorized access and securing data, both at rest and in motion. Several IoT access control models have been developed in the literature [76] , with cloud-assisted IoT access control models for AWS [77] , Google [24] , and Azure [78] . Traditional access control models are not adequate in addressing dynamic and evolving access control requirements in IoT.

Attribute-based access control (ABAC) [79, 80] , offers a flexible and dynamic access control model, which fits more into distributed IoT environments, such as smart home [81] , connected vehicles [71, 82] , and wearable IoT devices [6, 83] .

In addition to access control, communications in terms of data flow between various components in cloud-enabled IoT platform need to be secured from unauthorized data access and modifications. Thus, attribute-based communication control (ABCC) [84] A pertaining risk to these and other AI assisted system and applications is adversarial machine learning [86] using which adversaries compromise user data and privacy. In order to protect the data sets, differential privacy [72] can be applied to add noise. Cloud based medical data storage and the upfront challenges have been extensively addressed in the literature [87, 88] . Study [89] conducted semistructured interviews with fifteen people living in smart homes to learn about how they use their smart homes, and to understand their security and privacy concerns, expectations, and actions.

In future research, there are requirements to conduct interviews of practitioners to understand the security and privacy concerns while developing the smart hospital, and need to apply similar approach involving community residents, infrastructure manufacturers and stakeholders to develop other components of the smart connected ecosystem. Privacy preserving deep learning approaches such as collaborative deep learning or federated deep learning also need to be explored to train and deploy local models at the edge devices.

Within a connected ecosystem, users are constantly interacting with numerous smart devices and applications. One of main challenges in such an environment with billions of smart devices is performance 47 https://spectrum.ieee.org/telecom/security/tracking-covid19-with-the-iot-may-put-your-privacy-at-risk efficiency and quality of service (QoS 

IoT is an emerging technology that is being adopted by several nations across the world. One reason for the lack of constitutions and policies may be because the IoT differs from other network technologies and there is a lack of specific IoT standards. The research on constitution and policy, including engagement on public policy development debates, and IoT standards is necessary to successfully integrate privacy, accuracy, property and accessibility solutions in the smart communities.

To develop effective constitutional policies and standards, collaboration across governmental and nongovernmental organizations and industry partners as cloud providers, IoT manufacturers, would be beneficial.

Enabling a smart community requires thousands of low-power and low cost embedded devices together with large scale data analytics and applications. There are several implementation challenges involved in developing such large scale smart infrastructure. Fault tolerance and resilience are the challenges for reliable delivery of sensor data from smart devices to distributed cloud service.

Various failures can occur including face recognition, community infrastructure management, and 48 http://smartcities.gov.in/content/innerpage/guidelines.php 49 https://www.transportation.gov/smartcity 50 https://www.govtech.com/opinion/If-Only-One-US-City-Wins-the-Smart-City-Race-the-Whole-Nation-Loses.html emergency response in smart infrastructure. GEOgraphically Correlated Resilient Overlay Networks (GeoCRON) [95] is developed to capture the localized nature of community IoT deployments in the context of small failures. Research in [96] proposed a new fault-tolerant routing technique for hierarchical sensor networks. Another challenge for constant running and managing these IoT devices is costs related to energy , communication, computation, infrastructure, and operation. There is generally a tradeoff between benefit and cost for IoT applications [97] , however in the scenario of COVID-19 pandemic, expected benefits (saving lives, economic growth) should outweigh the operational and deployment costs.

Another challenge, for instance, various IoT devices communicate to each other or server to build a prediction model. If the local model is not able to predict accurately due to data duplication or other reasons, there will be no point to build such a model. This study [98] proposed a game theoretical analysis to allocate more storage capacity in a cost-effective manner, which achieves to maximize the benefits.

For the future directions, game theoretical approach can be used to analyze the smart infrastructures in terms of cost-benefit analysis. Furthermore, interdisciplinary research collaboration is inevitable to implement a smart connected ecosystem. There are several areas of research and engineering aspects, as machine-to-machine technology, artificial intelligence, deep and machine learning, predictive analysis, security and privacy, and others need to be merged and collaborative research approach is necessary in implementing, deploying, and managing a smart connected ecosystem.

IoT generate tremendous amount of data collected by physical devices, and this raw data is converted into valuable knowledge using AI and machine learning technologies. The "6V" (Value, Velocity, Volume, Variety, Variability, and Veracity) Big Data challenges for IoT applications are discussed in [99, 100] . The volume of data from IoT devices overwhelms storage capacities. There is not only storage issue, but the data needs to be organized properly so that it can be retrieved and processed in a timely manner. Data duplication is a data storage issue when an organization has multiple copies of the same data source. For example, a user has multiple wearable smart bracelets (smart hospital bracelet, smart grocery bracelet, and RFID antibody tag bracelet), these wearable devices will collect similar kind of data from a user which can create an issue of data duplication.

Machine Learning (ML) based applications require a large amount of valuable data for correct prediction, however, complicated and insufficient data can be an issue to the accuracy of the learning and predictive models. In addition, ML approaches need further research and development to deal with such heterogeneous and constantly evolving sensory data inputs. For instance, a locality-based COVID-19 patient detection model based on early symptoms learns with the collaboration with smart nursing home data sets and smart child care data sets. The prediction model can be biased towards elderly people if the number of patients in smart nursing home are more than smart child care. To overcome this problem, both models can learn at their edge networks using collaborative deep learning [101] . Research on these open challenges will help early development and deployment of future smart communities.

In this paper, we propose future smart connected community scenarios, which are blueprints to develop smart and intelligent infrastructures against COVID-19 and stop similar pandemic situations in the future. The autonomous operation with low human intervention in smart communities enable safe environment and enforce preventive measures for controlling the spread of infection in communities.

Data-driven and AI assisted applications facilitate increased testing, monitoring and tracking of COVID-19 patients, and help to enforce social distance measure, predict possible infections based on symptoms and human activities, optimize the delivery of essential services and resources in a swift and efficient manner. The paper discussed different use case scenarios to reflect smart applications and ecosystem.

The plethora of IoT devices enable huge data collection in different sectors including healthcare, home, supply chain management, transportation, environment, and city, which raises user concerns. In addition, the implementation of proposed smart connected scenarios face other challenges including legislation and policy, deployment cost, interoperability etc. which have also been discussed in the paper. We believe that our vision of smart communities will ignite interdisciplinary research and development of connected ecosystem to prepare our world for future such outbreaks.

Smart items, fog and cloud computing as enablers of servitization in healthcare

Enabling health monitoring as a service in the cloud

Opportunities and challenges of the internet of things for healthcare: Systems engineering perspective

Security and Privacy in Smart Farming: Challenges and Opportunities

Cloud-assisted industrial internet of things (IIoT)-enabled framework for health monitoring

An access control framework for cloud-enabled wearable internet of things

The case for vm-based cloudlets in mobile computing

The virtual object as a major element of the internet of things: a survey

An IoT-aware architecture for smart healthcare systems

Enhancing the quality of life through wearable technology

An IoT-Aware Architecture for Smart Healthcare Systems

Fall detection -principles and methods

Flexible technologies and smart clothing for citizen medicine, home healthcare, and disease prevention

Privacy-Preserving Deep Learning

A health smart home system to report incidents for disabled people

Correlation between real and simulated data of the activity of the elderly person living independently in a health smart home

Study and implementation of a network point health smart home electrocardiographic

Model and Simulator of the activity of the elderly person in a Health Smart Home

Health smart home -towards an assistant tool for automatic assessment of the dependence of elders

Detecting Health and Behavior Change by Analyzing Smart Home Sensor Data

Patient status monitoring for smart home healthcare

Automated Cognitive Health Assessment From Smart Home-Based Behavior Data

Towards a Distributed Estimator in Smart Home Environment

Access control model for google cloud iot

A Closer Look into Privacy and Security of Chromecast Multimedia Cloud Communications

Investigating Security and Privacy of a Cloud-based Wireless IP Camera: NetCam

A Testbed for Privacy and Security of IoT Devices

An Experimental Framework for Investigating Security and Privacy of IoT Devices

Analysis of IoT Traffic using HTTP Proxy

Convergence of MANET and WSN in IoT urban scenarios

A method to make accurate inventory of smart meters in multi-tags group-reading environment

Smart Spare Part Inventory Management System with Sensor Data Updating

An IoT Application for Inventory Management with a Self-Adaptive Decision Model

An IoT Based Inventory System for High Value Laboratory Equipment

IoT Applications on Secure Smart Shopping System

Aeon: A Smart Medicine Delivery and Inventory System for Cebu City Government's Long Life Medical Assistance Program

Study of smart warehouse management system based on the IoT

Iot based grocery management system: Smart refrigerator and smart cabinet

Short on Time? Context-Aware Shopping Lists to the Rescue: an Experimental Evaluation of a Smart Shopping Cart

The Smart Shopping Basket Based on IoT Applications

Sysmart indoor services: A system of smart and connected supermarkets

Enabling rfid in retail

Secure V2V and V2I Communication in Intelligent Transportation using Cloudlets

Enabling and Enforcing Social Distancing Measures using Smart City and ITS Infrastructures: A COVID-19 Use Case

Smart community: an internet of things application

Internet of Things and Big Data Analytics for Smart and Connected Communities

Smart Cities, Big Data, and Communities: Reasoning From the Viewpoint of Attractors

A Community-Based IoT Personalized Wireless Healthcare Solution Trial

An integrated cloud-based smart home management system with community hierarchy

Modeling 'Thriving Communities' using a Systems Architecture to Improve Smart Cities Technology Approaches

Research on Key Technology for Data Storage in Smart Community Based on Big Data

A use case in cybersecurity based in blockchain to deal with the security and privacy of citizens and smart cities cyberinfrastructures

Internet of Things for Smart Cities

Understanding Smart Cities: An Integrative Framework

An Information Framework for Creating a Smart City Through Internet of Things

Foundations for Smarter Cities

Long-range communications in unlicensed bands: the rising stars in the IoT and smart city scenarios

Smart health: A context-aware health paradigm within smart cities

Smarter Cities and Their Innovation Challenges

Everything you wanted to know about smart cities: The Internet of things is the backbone

UAV-Enabled Intelligent Transportation Systems for the Smart City: Applications and Challenges

A Collaborative Mechanism for Private Data Publication in Smart Cities

Scalable Mobile Sensing for Smart Cities: The MUSANet Experience

An introduction to multi-sensor data fusion

Smart cars on smart roads: problems of control

The security and privacy of smart vehicles

Real-time object detection for "smart" vehicles

Predictive Active Steering Control for Autonomous Vehicle Systems

Collision Avoidance and Stabilization for Autonomous Vehicles in Emergency Scenarios

Learning driving styles for autonomous vehicles from demonstration

Dynamic groups and attribute-based access control for next-generation smart cars

The algorithmic foundations of differential privacy

On the security and privacy of Internet of Things architectures and systems

Privacy through pseudonymity in user-adaptive systems

Privacy protection: p-sensitive k-anonymity property

Anas Abou Elkalam, and Abdellah Ait Ouahman. Access control in the Internet of Things: Big challenges and new opportunities

Access control model for aws internet of things

PARBAC: Priority-Attribute-Based RBAC Model for Azure IoT Cloud

Guide to attribute based access control (abac) definition and considerations (draft). NIST special publication

A unified attribute-based access control model covering DAC, MAC and RBAC

Authorizations in cloud-based Internet of Things: current trends and use cases

Authorization framework for secure cloud assisted connected cars and vehicular internet of things

Poster: IoT SENTINEL-An ABAC Approach Against Cyber-Warfare In Organizations

ABAC-CC: Attribute-Based Access Control and Communication Control for Internet of Things

Iot passport: a blockchain-based trust framework for collaborative internet-of-things

Adversarial attacks on medical machine learning

Development of private cloud storage for medical image research data

Extensive medical data storage with prominent symmetric algorithms on clouda protected framework

End user security and privacy concerns with smart homes

MIFaaS: A mobile-IoT-federation-asa-service model for dynamic cooperation of IoT cloud providers

Internet of Things-New security and privacy challenges

A framework for Internet of Things-enabled smart government: A case of IoT cybersecurity policies and use cases in US federal government

The United Kingdom's Emerging Internet of Things (IoT) Policy Landscape. Tanczer, LM, Brass, I

The United Kingdom's Emerging Internet of Things (IoT) Policy Landscape

The Internet of Things (IoT) and its impact on individual privacy: An Australian perspective

Resilient overlays for IoTbased community infrastructure communications

Enabling reliable and resilient IoT based smart city applications

Cost-Benefit Analysis at Runtime for Self-adaptive Systems Applied to an Internet of Things Application

Cost-benefit analysis game for efficient storage allocation in cloud-centric internet of things systems: a game theoretic perspective

Internet of things: Vision, future directions and opportunities

Deep learning for IoT big data and streaming analytics: A survey

Learner's Dilemma: IoT Devices Training Strategies in Collaborative Deep Learning