Questioning if Macro-Systemic Top-Down Approaches Should Be Used to Develop Smart Cities of the Future

Cities differ greatly from one to another, even if they serve the same essential purposes. Moreover, they should collaborate at all levels for producing synergistic effects, therefore, their major smart solutions should be standardized or at least compatible, whilst adapted to the local contexts. Cities are systems of systems and contain a large variety of components, each of them presenting different characteristics. Given the vast range of technological choices and the unforgiving overall pressure, governors struggle to implement best specific smart solutions for their cities, by usually adopting bottom-up methods, thus, endangering future possible compatibilities with other local or international systems. This research stresses the importance of top-down smart city deployments, by acknowledging also the local specificities. It investigates the status of relevant international contributions, questioning what kind of approaches should be applied by public administrations. The results point out to complex challenges that exist by revealing the status of standardization proceedings, the authors’ perspective on the enabling technologies and their dimensions, as well as a related case study concerning a local community of Romania so as to bring the beforementioned issues to light. The main objective of this research is to highlight the associated problematics and draw attention to the urgent need of a new macro-systemic top-down attitude for developing smart cities.


Introduction
There are many complex challenges for developing smart cities of the future and they depend on a substantial variety of factors, e.g. economy, sociology, engineering, management, governance. As reflected in Fig. 1, smart cities must be built, organized, and coordinated in such a way that the overall status-quo, legislation, technology, solution feasibility, and community's acceptance overlap to the highest extent possible, by finding common denominators for reaching greatest positive effects with smallest efforts. The objective of this paper is to investigate the approach required by cities to successfully implement the best smart solutions in conjunction with their local communities in order to benefit from the right enabling technologies, standardized or compatible systems, and to be able to interconnect with other cities for producing the needed synergistic effects.

Literature Review and Problematics
This research analyzed the status and evolution trends of smart city systemic elements; mainly digital, to include them in the beforementioned investigation.

Sensors
Smart city performance and adaptability rely on an interconnected ocean of sensors that are already partially present in the cities of today, which are seen by some authors as "senseable cities" (Ratti and Claudel, 2016, Part III). Accelerometer, temperature, light, micro-electro-mechanical systems, and switch sensors are indispensable in our current applications-e.g. see Fig. 10-, but there are many other types available: barometer, GPS, gyroscope, magnetometer, proximity indicator etc. (Sethi and Sarangi, systems operate at desired performance level including the critical situations, such as emergencies and disasters. These sensors are complemented by humans' direct feedback, which will probably never be completely substituted by any other automatic mechanism in the following decades. Today's cities are responsible for consuming and manipulating most of world's resources, therefore, they need to implement greener and more efficient day by day procedures, with the help of their sensors specialized in infrastructure, buildings, electricity, water distribution, transportation, healthcare, services, and surveillance etc. (Hancke and Hancke Jr, 2012). As a concrete example, from the public space point of view, Barcelona's Sentilo platform uses more than 1800 sensors to monitor data about energy, noise, urban lab, garbage collection, and parking spots, and this number is expected to increase over time (Sinaeepourfard et al., 2016). Many sensing applications are dedicated to energy efficiency and environmental protection; as these topics are top concerns of the twentyfirst century (see, for instance, Pradeep Kumar Sensors have evolved towards a multitude of forms and technologies-including nanotechnology-while adding the new versatile mobility dimension, as in the case of mobile phones, which transform the cities into largely covered, dynamic and interactive entities (Hancke and Hancke Jr, 2012, pp. 395-396). Three sensing platforms are to be highlighted for the ubiquitous coverage: wireless network sensors, IoT systems, and Cloud of Things [ibid, pp. 399-403]. In the case of no existing energy and communications infrastructure, e.g. postdisaster situations, there are already autonomous solutions which acquire and transmit necessary data to the decision makers (Habibzadeh et al., 2017).
Virtual sensors are a distinct important category, which allow simultaneous readings by multiple users in a distributed system. They are simple-to-use programming abstractions of real sensors, which exist locally or distantly-e.g. in the Cloud (Vermesan and Friess, 2014, Chap. 3.7.4)and make the implementation of the Sensingas-a-Service paradigm possible (Vermesan and Friess, 2015, Chap. 8).

Data
Data collected by sensors are stored, aggregated, processed, distributed, correlated, analyzed, and interpreted in certain schemes, to support human or automatic collaborative decision-making and needed outcomes. These data may be structured-i.e. data is recorded in files with fixed input fields-or unstructured-photos, videos, social media, emails etc. The challenges in managing the data explosion shown in Fig. 2 are multiple, such as ultra-fast acquisition of data, reliable large storage capacity for all circumstances, filtering, prioritization, processing performance, general accuracy, models for interpretation, and information security. Data centers, big data analytics, and cloud computing are few examples of technologies that mostly manage these challenges.

Fig. 2: Estimation of data explosion
Source: Odini (2016) Data have three main characteristics-volume, velocity, and variety-, they produce value chains within the processes, and trigger human or automatic actions.
Some of main sources for generating data are sensors, smart mobile devices, and web services (Sinaeepourfard et al., 2016). Each of them generates large amounts of data, for example Barcelona's Sentilo platform registers beyond 1.3 million records daily in its data bases [ibid]. These big data are used for predictive analysis, decision automation, data driven businesses, and business performance management (Hashem et al., 2016, Fig. 3), whilst they are systematically processed by smart analyzers, which split, map, shuffle, sort, reduce and deliver all the ingurgitated amounts from the environment [ibid]. To these sub-processes, Khan et al. (2015) add the integration, visualization, as well as supervised and unsupervised learning.
Importantly to underline is that these processing activities happen not only on fixed platforms, but together with mobile platforms which involve smart mobile devices, fogs and clouds (Arfat et al., 2017). To assure interoperability and hide the hardware complexity of many public and private devices operating within a city, more abstraction layers are usually required for the development of applications. For example, iSapiens platform suggests four layers: data, information, knowledge, and wisdom layers (Cicirelli et al., 2017).
Open data, supported by initiatives such as the Open Government Partnership launched in 2011, enable the participation of many actors to faster creation of hard and soft solutions, innovation, synergies, knowledge sharing, and other very positive effects. Last years have imposed new approaches that encourage an active participation of the society, such as crowd sourcing, gamification, co-creation, self-organization, and others (Dijk and Teuben, 2015). For example, Helsinki made available 1030 data bases to the public in 2013, addressing many aspects of the city life, so as to be valorized with these new approaches (Hashem et al., 2016). Moreover, the publishing of these data, facts, and trend lines tempers the population from being suspicious and arguing groundlessly, thus, it promotes a trustful and pragmatic cooperation environment (Goldsmith and Crawford, 2014). Additionally, data democratization and digital access strengthen the social equity (Nguyen and Boundy, 2017), and promote the development of new governance models for smart cities as well (Barns, 2018).

Networks
Many years ago, scholars have indicated that the evolution of cities is closely linked to telecommunications infrastructure and Fig. 3 shows the pressure on the requirements for 5G communication (Vermesan and Friess, 2014).

Fig. 3: 5G requirements
Source: Vermesan and Friess (2014) Networks link people to people, people to machines, and machines to machines in all activity areas: businesses, healthcare, education, governance, public utilities etc. To interconnect this large diversity of entities, standardized platforms must be created and put into the market, by regulating hardware, software, and protocol solutions. Worthy and important to be mentioned is the networking of cities, as well as the networking of services through the Anything-as-a-Service platformsaka.

Processors
Generally, vendors struggle to produce processors and microcontrollers which are multi-core, fast, versatile, reliable, small, cheap, easy to program, and able to process large quantities of data to incorporate new technologies-e.g. for face recognition in dynamic environments, and to synchronize with other processors and systems.  Fig. 1). According to some scholars, the balance between these two last computing approaches should be thoughtfully addressed and designed (Garcia Lopez et al., 2015).
In certain cases, it is necessary to locally learn from remote devices and users, recognize patterns, and apply inter-device controlling, therefore, the processors must securely communicate, share data and other resources, as well as make use of remote computing power (IEEE Staff, 2016).

Software and Applications
Software populates and commands almost all intelligent machines on the planet. It may be of various types, structured on different layers, but the main aspect is that it makes the devices "alive", so that they act and react, learn, communicate, and collaborate. Software can be updated; therefore, the devices can be cheaply "reinvented" by keeping the same hardware.
There are different types of software applications, depending on their purpose, capabilities, and localization. For example, some applications are functionally limited, embedded in microcontrollers and perform edge-computing tasks (Garcia Lopez et al., 2015). Some applications are versatile and used locally in larger ICT systems, whilst others "live" in local micro-clouds or the Cloud and are triggered from distance as Software-as-a-Service (Gubbi et al., 2013, Sneps-Sneppe and Namiot, 2016). It is important to do the right job at the right time, in conjunction with the hardware, including the critical environments.
In smart cities, high-level user-oriented applications should cover all activity fields (e.g. transportation, healthcare, energy management), by integrating all necessary enabling technologies, such as Cloud, Fog, IoT A special software is the middleware, which is a virtualization bridge between the hardware-specific layer which widely differ from one vendor or technology to another, and the end-purpose applications which are the main target for the developers who want to make the devices intelligent so as to fulfill their tasks. To do that, today's middleware has to offer the developers smart APIs which must be versatile, simple to use, and can enable fast deployments. By using APIs, the integration of devices, networks and applications is facilitated. It is also worth to mention the abstract digital virtualization layers that operate in multiple IoT architectures and platforms-e.g. FIWARE-, as well as the smart cities dedicated operating systems which integrate IoT solutions and various public services (Vermesan and Friess, 2015), but they are far from being mature enough or even not to exist in many cases. A concrete example of an organization that adopted such a vision and solution is the City Protocol Society. Based on its own functional platform and the Open & Agile Smart Cities initiative (OASC), it develops a network of more than 50 smart cities using  Finally, a special subject to be highlighted in this subchapter is the Artificial Intelligence-AI-(Nilsson, 2014, Pan, 2016), which evolves based on a special kind of software and does special actions like perceiving, interpreting, communicating, learning, making decisions, and triggering actuations in the real world. AI has already started to be used also in the smart cities (see, for instance, Mathur and Modani, 2016), in various fields, e.g. transportation (Agarwal et al., 2015), public and private buildings (Adio-Moses and Asaolu, 2016, Skouby and Lynggaard, 2014), energy (Mrówczyńska et al., 2014) etc. Great successes are yet to come in this area, although one must say they induce definite risks to humans, therefore, they must be managed carefully.

Actuators
Actuators change the status of the environment by influencing its components and/or parameters (Sethi and Sarangi, 2017). They are devices powered by any type of energy and vary from very simple ones-e.g. on-off switches (see Fig. 10)-to very complex-e.g. control systems of robots (see, for instance, Vo et al., 2018). One could further extend this standard definition by including AI, machines, and even humans that trigger actions to satisfy requests or necessities, regardless whether they are in the physical or virtual reality.
Like in the case of virtual sensors, virtual actuators control a set of real-life actuators, being triggered by stimuli of the physical or virtual world. This could imply the usage of dedicated middleware to manage and facilitate the interaction between the hardware itself and the higher level services (see, for instance, Merlino et al., 2015).

Electric Energy
Anything in this world works by using energy. Its production, transmission, distribution, storage, and usage critically influence the entire environment. Humankind has acknowledged this very tight influence, sometimes with negative impact, therefore, all energy subsectors must become smart. "Smart energy aims at greener energy generation, lower energy consumption, an energy consumption pattern with flattened peaks, and a resilient distribution grid" (Dijk and Teuben, 2015).
To implement this vision, following approaches and instruments should be mixed and harmonized to the highest extent possible within the city premises: distributed generation with renewable sources, smart grids, microgrids, smart metering, lower usage through gamification, responsive devices, seasonal thermal energy sources, excess heat use, electric vehicle charging, power of combining, and cooperation in energy markets (Dijk and Teuben, 2015). These aspects apply to all areas of smart cities: buildings, green spaces, mobility, healthcare, public services etc. A vast international literature addresses all these aspects, in conjunction with the smart city topics or not, as smart energy is one of the most pressing themes at worldwide level (see, for instance, Connolly  In the new Energy Cloud concept introduced in 2015 by Navigant Research (see Fig. 5), IoT plays a central role and constitutes a framework for "clean, distributed, and intelligent energy systems. IoT technologies are well-suited for integrating the two-way flow of power, the intermittency of wind, and solar, and behind-the-meter processes such as onsite storage or EV integration" (Strother and Lawrence, 2016).  Another term to describe the energy management system of the future is the "Internet of Energy" (see Fig. 6), which implies a "networked Smart Grid that is largely automated, by applying greater intelligence to operate, enforce policies, monitor, and self-heal when necessary. This requires the integration and interfacing of the power grid to the network of data represented by the Internet, embracing energy generation, transmission, delivery, substations, distribution control, metering and billing, diagnostics, and information systems to work seamlessly and consistently" (Vermesan and Friess, 2014).  To be highlighted herein are the energyrelated research directions for autonomy, production self-capacity, conversion, trading platforms, automatic purchase, optimum usage, standardization of new technologies, harvesting of mechanical, thermal, radiant, and chemical energy, along with general power management [ibid, Chap. 3.9], as well as the issues and effects of deploying largescale energy systems managed by ICT infrastructure (Song et al., 2011).
Last aspect to be mentioned is the progressive introduction-with good resultsof AI in the management of all energy subfields (see, for instance, Mrówczyńska et al., 2014).

Human Factor
Humans are very special actuators of the environment and cities. . Smart technical solutions must support today's citizens, but they are unpredictable, reluctant, and not skillful in all domains, as well described by Alexander et al. (2016). Therefore, technical solutions should be error proofed and as user friendly as possible, even if information, options, devices, and systems are more and more abundant and complex.
To increase the prosperity of cities in every sense, all actors should be constructively involved-desirably by digital platforms (Goldsmith and Crawford, 2014)-, including methods like civil hacking (Townsend, 2013). New ways to attract, train, involve, and change behavior of humans are the innovative methods such as gamification, crowd sourcing, co-creation etc., which produce better and better results every year (Dijk and Teuben, 2015). The involvement will evolve over time towards e-democracy, which is based on principles like e-community, egovernance, e-budgeting, e-policy making, e- People-to-machine interfaces are very important to make life easier and more efficient. They may be of many types, ranging from wearables to stand-alone instruments, and developed by various interaction technologies-voice, video, touch, keyboard etc.

Platforms, Models, Processes, Standards
In all activity fields, there are numerous and ever-changing visions, policies, strategies, methodologies, use cases, models, platforms, and standards that are available or being imposed at different levels. For example, Fig.  7 indicates today's tendency to transform the ICT network paradigm. It is clear there will always be a diversity of approaches from one community to another, from one city to another, from one state to another, depending on their general status, culture, population, business needs, and governmental maturity. This diversity is produced by multiple top-down and bottomup development approaches, depending on a mix of influencing factors. But there must be both a "common language" and solutions to interconnect all implemented systems, to facilitate collaboration and prosperity. Unfortunately, common solutions could sometimes mean bigger risks, e.g. concurrent hacking threats (see, for instance, Vermesan and Friess, 2015, Chap. 6).
In the meantime, standards of international and national bodies evolve in all domains and cover architectures, data management, activity flows, non-functional aspects etc. Their benefits are obvious, if they are correctly applied (Trivedi, 2016). They reduce risks and invested money, determine faster integration and synergies with similar entities, and allow simpler maintenance etc. It is self-understood that these standards produce diverse results from one case to another in terms of aesthetics, features, and performances etc., but finally the outcomes should be mostly compatible. and valorize, configurable, autonomic, scalable, upgradeable etc. From processes point of view, these platforms may be much different from one organization to another, as they are defined and controlled by humans, but-in any case-they have to be adapted to the new trends, e.g. to become mobile, datacentric (Dijk and Teuben, 2015), adaptive, event-driven, highly distributed, as well as capable of dealing with unreliable data/resources or working in volatile environments (Vermesan and Friess, 2014, Chap. 3.6). At the same time, these new principles should be standardized and framed, e.g. as Fig. 8 suggests for IoT landscape. Of course, there will always be front runners who will impose new innovative trends at large scale-e.g. digital economy, shift to subscription-based business models, shift from possession to use, shorter product life cycle etc.-and followers who will have to adapt to these new trends (Dijk and Teuben, 2015).

Systems and System Integration
Cities can be automated and improved almost indefinitely within the limits of available resources and community interests, while maintaining respect for the environment, tradition, and posterity.
Depending on the field of activity and usage, future smart city systems must provide open data and be global and inclusive by connecting people, businesses, machines, knowledge, and other cities as well. One of the Main objectives is to have comprehensive cyber-physical highly-intelligent systems to cover most of the needs and opportunities, by integrating data, sensors, networks, processors, applications, actuators, processes, business models, energy sources, and humans. More and more, future systems will also integrate distributed intelligence, software-defined communications, multipurpose networks, cloud/fog/cloudlet computing, data analytics, and batterypowered edge devices. All these require competent planning, synchronization, and management (Beurden et   Efforts are done to frame the technology, e.g. as for the above-mentioned Network of Things model developed by NIST (Strother and Lawrence, 2016), which defines five elements that compose the IoT systems: sensors, communication channels, aggregators, Cloud, decision triggers, and external utilities (see Fig. 10). The ability to innovate and bring together individual solutions into aggregates-e.g. public lamp posts with integrated facilities for telecommunications, security, traffic management, and ecologic management-will significantly reduce the cost, space, and speed of implementation (see Fig. 11).

Systems Security, Safety, and Trust
Today's risks associated with terrorism and cyber-attacks of different natures (see Fig.  12) determine the necessity to better secure the mechanisms that operate in smart cities. Systemic vulnerabilities related to infrastructure-energy, water supply, healthcare, telecommunications, transportation, etc.-, economy and social services are not permitted and can lead to chaos in few moments. This is the perception of city residents too, as highlighted by a significant survey realized by Osborne Clarke (2015). The negative effects could be multiplied, if we suppose that citizens lose a part of their ability to adapt and react rapidly in the future because of the automated systems which they become accustomed to. Therefore, all security related aspects must be treated with extreme caution.  Solutions to counteract the cyber-security threats are already in focus since many years, but smart city systems become more and more complex since backdoors and breaching niches conclusively multiply and may be exploited by hackers (see, for instance, Ferraz et al., 2016, International Electrotechnical Commission, 2016, International Telecommunication Union, 2015). All activity fields and areas are targeted and must be safeguarded, e.g. economy (Kumar, 2016), healthcare (Zhang and Shen, 2015), and society categories (Taylor and Rooney, 2016).
From the safety point of view, smart city systems must solve all aspects related to the misuse of public infrastructure and services by humans and machines. At the same time, they have to predict, to be resilient and protective in case of natural/artificial disasters or critical failures (Arafah and Winarso, 2017). These capabilities are not easy to implement, especially that they imply the consideration of many scenarios, as well as additional hardware and/or software.
Concerning privacy, personal and business information must be protected according to user's preferences, e.g. identity, location, and digital operations etc. For this purpose, advances are required with respect to the improvement of cryptographic techniques, privacy-to-design concepts, self-configuring access control, and others (Vermesan and Friess, 2014, Chap. 3.8.3).
All aspects above induce trust or mistrust related to the technologies and solutions that are deployed in smart cities, by guaranteeing their success on the long run or not. Therefore, they must be treated carefully, starting with the requirements engineering phase.

Local Context and Society
The overall local context-e.g. culture, legislation, resources, political status-is very different from one city of the world to another. Consequently, the success of

Theory and Research Directions
Given the fast and ever changing social, economic, territorial, political, and natural environments, a mandatory transversal feature of future smart city systems is the capacity to learn, adapt, and respond in real time, irrespective of the implementation solutions. The argument is that, at present, scientists have not yet identified and mastered all algorithms that rule communities and nature. These systems must be capable of permanently monitoring and identifying patterns that exist in the ecosystems and that cannot be framed by simple mathematical models. Over time, they will be able to determine sophisticated algorithms that govern the real world and then they will improve their performance accordingly. Such adaptive systems rely on data accumulation, in absolute and relative terms, targeting various phenomena trends, seemingly chaotic facts-but which could be characterized by fractal patterns for example-, synergetic effects of overlapping causes, human behavior, unstructured feedback of users etc. A consequence of such adapting capabilities is that, depending on the local conditions, an identical smart system could evolve and behave very differently from one world city to another.
To deploy attractive, efficient, effective, and trustful smart city systems, as well as to interconnect these systems to each other and to the systems of other smart cities, compatible solutions must be adopted and implemented. This requires right vision, planning, requirements engineering, implementation, field deployment, operation, maintenance, and upgrades. Therefore, the approach should mostly be top-down by making use of international/national standards and regulations as grounding platform.
In theory, it is possible to standardize the inputs, outputs, interfaces, characteristics, business models, and processes of smart city systems, by covering main aspects concerned: technical, managerial, materials, equipment, and information exchange. It is obvious that all details of smart cities can neither be standardized nor implemented the same way, given the specific local status, resources, existing systems, and politics.
Still, it is essential to urgently create international frameworks for consolidating the future smart cities with as straightforward solutions as possible. Cities are already finding and implementing their own solutions for facing their challenges and therefore, if not providing these frameworks, they will become much different during next decades, consequently, reducing the multiple synergetic effects that could be generated. Therefore, the existent, developing and foreseen smart city standards must be actively acknowledged and promoted (i.e. first research direction of this paper). Even if these international frameworks are acknowledged by city administrations, it is furthermore necessary to understand their sense and implications, as well as their related enabling technologies (i.e. second research direction of this paper), which must be synthesized and brought to light by scientific studies.

Journal of Administrative Sciences and Technology
It is also to be highlighted that regardless of the city business models themselves, smart solutions must be firstly adapted to and validated by the local community and direct stakeholders (i.e. third research direction of this paper), to meet the real expectations and needs, on top of gathering all necessary support for the upcoming efforts. The related results could be surprising, for example some seemingly "low hanging fruits" solutions could encounter multiple blockages, while some tougher-to-implement solutions-e.g. general platforms with middleware and virtualization-could enthusiastically boost the development of the city on the long run. So, future acceptance and implementation feasibility must be-in the first place-very well evaluated, planned, and introduced by pilot experimental projects, in a collaborative approach.
Finally, it is worth suggesting that in our competitive and challenging world, by applying the vision of smartly interconnecting the cities at all levels (i.e. forth research direction of this paper), they would synergistically work together to benefit from opportunities, solve necessities, overcome problems, handle threats, and learn faster from each other. So, this desideratum should be actively supported, researched, and implemented.  Besides their experience, the authors reviewed the international specialty literature and surveyed 332 persons from Timis County (Romania), so as to determine the required outputs of the research, which are highlighted in the following two chapters, together with their subchapters.

Perspective of Standardization
International standards were published with respect to smart cities and their elements, but not all parts of the puzzle exist, so as to completely shape the future smart cities as systems of systems. Important efforts have been made by international bodies and, in parallel, by the European Commission, CEN, CENELEC, ETSI, and national bodies from Standardization-ISO-in its "Smart cities: Preliminary report 2014" (ISO, 2015). This report underlines the ongoing hard process of synchronizing the international and national bodies which develop concomitantly their standards of which some are being defined for specialty niches and others are being more generally overlapped by parallel definitions. Table 1 indicates up-to-date highlights for the work of ISO, ITU, IEC, and IEEE on subjects that are relevant to the systems and activity fields of smart city-e.g. ICT, electronics, and energy. • IoT: reference architecture; use cases; interoperability for IoT systems; • Sensor networks: services and interfaces, reference architecture; sensor network and its interface to smart grid system, generic sensor network application interface; • Underwater acoustic sensor network.
ITU ITU (2019a) -L series: four sections related to key performance indicators for smart sustainable cities.

ITU (2019a) -Y series, sections:
• Next generation networks: frameworks and functional architecture models; service capabilities and service architecture; interoperability of services and networks in NGN; • IoT and smart cities & communities: infrastructure, connectivity, and networks; frameworks, architectures, and protocols; services, applications, computation, and data processing; management, control, and performance; identification and security etc.; • All supplements: guide for city leaders, master-plan, integrated management, multi-service infrastructure, ICT architecture, stakeholders' engagement etc. • Applications of wireless sensor networks in next generation networks; • ICT, environment, and climate change; • Internet of things (IoT) and machine-to-machine communication (M2M); • United 4 Smart Sustainable Cities; • Trust in ICT.
IEC IEC (2019a) -Sections related to IEC's specific work on: connecting things, smart energy, mobility, water, smart buildings, and city services.
IEC (2019b) -Section related to projects and publications: city service continuity against disasters -the role of the electrical supply; smart cities reference architecture methodology. (2017)  Some people acknowledge and underline that the standardization for different parts of smart city systems is missing and have made related gap analyses, e.g. ISO (ISO, 2015, Chap. 8), and ITU (ITU, 2015, Chap. 7), whicheven if published some years ago-are mostly valid, as research results and global consensus are slowly reached by international standardization activities. Same scholars confirm also these existing gaps (see, for instance, Bibri and Krogstie, 2017, Table  2). The conclusion is that there are still many aspects to cover and gaps to fill-out for completing the overall landscape of smart city standards. At present, one of the most important aspect is probably acknowledging and highlighting the smart city as a system of systems and creating the related regulations. In this sense, Fig. 14 shows the abstraction level for the work of some standardization bodies.

Perspective of Technological Solutions
There are many solutions proposed by experts of various fields, communities, public institutions, and private companies, as there are very different realities, needs, and opportunities for cities which decided to become smarter. Under the reality pressure, many of the cities have already adopted and implemented different solutions. Today's main approach is bottom-up, in the sense that each city finds its own way to upgrade itself by sometimes inspiring themselves from other cities and/or by adopting at-hand solutions proposed by private companies which mostly impose-as large stakeholdersthe technological trends and solutions in a top-down manner. The worldwide leading companies offering smart city solutions are benchmarked in Fig. 15, by their strategy-i.e. vision, go-to-market strategy, partners, product strategy, and geographic reach-and by their execution capacity-market presence, sales & marketing, product performances & features, product integration, and staying power (Woods et al., 2017).  As previously indicated, unfortunately some smart city elements and functions are limitedly addressed by current standards, scientific literature, and solution suppliers. Significant examples are those related to the generic middleware, smart adaptive capabilities for all activity fields, and the instruments to smartly interconnect the cities by physical, virtual, and human platforms. This last aspect will probably be the next step to cover, after standardizing, the general concepts, systems, and implementation processes for stand-alone smart cities. Nonetheless, better results could be produced by considering the topic of smartly interconnecting the cities from the very first beginning of the standardization activities and, in parallel and in strong correlation, by the solutions suppliers.
Regarding the enabling technologies for smart cities and as already partly reflected by the previous chapters, the authors determined the following main ones, by analyzing and synthesizing a whole set of specialty literature: These technologies are most important for developing the smart cities of the future. Therefore, their role and valences must be acknowledged and used as such.

Perspective of Local Contexts
In accordance with the collaborative governance principles, 332 persons from Timis County of Romania were surveyed by the authors in July-August 2018 with respect to the 14 enabling technologies mentioned before. This process was applied for identifying local community's perception on future possible smart city solutions and, therefore, for their pre-validation.
The responders were categorized as below (see Fig. 16). Concerning their residency, 21% did not specify their exact location, while "rural" and "urban" mean localities outside Timisoara's metropolitan area. The survey results indicate that the responders' majority recommend all enabling technologies as being important for the smart development of Timis County localities (see Fig. 17):

Source: the authors
To be underlined in the survey is the response of public servants and dignitaries (see Fig. 18  Of course, all figures in the graphics from above are different from one local community to another and governors must take them carefully into consideration, if they want to be successful with the smart solutions which they are envisioning by themselves, but which could be greatly unsynchronized with the realities from the field. For example, in Timis County, strategies to introduce the enabling technologies in the public administration sector should consider a step by step approach, in order not to disturb the continuity of public activities and create unnecessary problems.
As a general conclusion for this subchapter, this should be further researched, local contexts could be very different from one world city to another, depending on a large variety of factors. This subchapter is only an exemplification. Therefore, to build cities which are smart in the real sense, public planners should firstly investigate the preparedness, perceptions, and positions of their local communities with respect to the enabling smart principles and technologies they want to apply. international economy, inclusive society, creative industries, social and intellectual capital, as well as urban sustainability.
It is well known that isolated cities may fall rapidly on a descendant slope, unless they undertake innovative measures to reconnect to the external realities (Tranos and Gertner, 2012, Chap. 4). Therefore, besides the upcoming inherent opportunities, interconnecting to other vibrant cities is a critical necessity for all cities, by paying special attention to the fact that the geography and typology are very different from one functional network of cities to another (see, for instance, Sigler and Martinus, 2017).
At the moment, the subject which is mostly treated with respect to the interconnecting of cities is the network governance (see, for instance, Martínez et al., 2016, Provan andKenis, 2008), so -at the moment -one cannot talk about "smart networking" (Vasile, 2018).

Discussion and Post-Research Questions
All fore-mentioned aspects prove that smart cities are very complex entities, composed of many systems that must operate in a smart and sustainable manner with care for humans and moral principles by using and enabling edge technologies.
A Question is which mandatory minimal smart components are to be implemented, integrated, and rolled out by using well streamlined phases, in order for the cities to be able to intelligently face their challenges, so that added value, services, quality of life, security, safety, cost reduction, and sustainability are at desired performance levels? (Q1).
Another question is to which extent smart city systems should be standardized (Q2), given that some international/national standards are already published or under rushed development?
An essential aspect is also that, until all needed backbone standards and regulations are ready, cities cannot wait. Cities are highly pressed by the increasing crowd of people and machines, political changes, globalization issues etc., not to mention the disruptive technologies-e.g. sharing economy, crypto-currency, AI, crowd sourcing, co-creation, XaaS-which emerge and create deep imbalances. Therefore, cities must invest money in solutions that are aligned to future trends and contextual developments. Hence, a natural question emerges, that is how much cities should/could wait for the appearance of standardized smart concepts, related legal decisions, and the concurrent technologies, including the succeeding implementation, field deployment, operational, managerial actions? (Q3) One conclusion of this research with respect to the international literature and standards indicates a complex multidimensional perspective on the development of smart cities. At this point, in the absence of fullyestablished standardized frameworks and guidelines, it is no wonder that governors are striving to break through the vastness of implementation alternatives-sometimes without understanding all related details and implications-and they are in difficulty to take right decisions. This is one of the reasons why the actual development of cities is bottom-up.
Nine major ICT-related dimensions come to light with respect to deploying smart city systems, which represent the synthetized standpoint of the authors. They are the following: First dimension (system type): Sixth dimension (process type): user-centric, data-centric, event-driven, adaptive, working in volatile environments etc.; Seventh dimension (user type): public administration staff, city citizens and humanbased organizations, machines-incl. vehicles, robots, and drones-, as well as connected cities; Eighth dimension (interconnectivity type): wired vs. wireless vs. mobile vs. iBeacons vs. machine-to-machine vs. IoT, subsystem to subsystem, system to system, system of one city to system of another city, city to city-i.e. smart complete networking; Ninth dimension (systemic characteristics): performance-incl. responsiveness, efficiency, and effectiveness-, quality-incl. reliability-, multifunctionality, interoperability-incl. interconnectivity-, complexity, modularity, scalability, adaptability-incl. configurability and non-obsolescence-, autonomy, resilience, transparency, data openness, coparticipation-incl. crowdsourcing-, security and privacy exposure, degree of automationincl. AI-, predictability etc.
By acknowledging all these dimensions, it is hard to propose a visionary model for assuring the best and shortest implementation path for transforming cities into smarter cities. The identification of a set of models to handle typical city situations could be the subject of another research, i.e. to indicate based on experience, experiments, and computed scores which are the best technological solutions, project setups, and roadmaps to develop the cities in an efficient and effective manner. Thereby, the components of the nine dimensions could be quantified in various axes-e.g. competence to implement/manage/maintain, budget to implement, time to implement, possibilities of integration to other standard systems and then assembled to create the models. The associated results of this research could also be used, besides the cities themselves, by the international/national standardization bodies to decide upon the priorities to define the most critical standards needed by the world, because if we consider the important topic of smart city applications, the review of international standards clearly shows that the needed elements, integration, and processes are only partially crystalized and framed at present.
Given this complexity and the fact that the overall status-quo-e.g. culture, legislation, resources, context is very different from one city of the world to another, it is hard to provide a straightforward and generally valid response for the three questions from above (Q1, Q2, Q3). All outlined aspects make the idea that each smart city must tailor and speed up their smart transformation by themselves plausible, by following for compatibility and intra-/interconnectivity reasons, as much as possible, the related international standards and by letting the door open for future upgrades, recalibrations, reconfigurations, adaptations, integration of new systems, and smart collaboration with other cities.
A synthetized response for Q1 is that the minimal smart components for a rapid transformation and optimum functional level should be found by a thorough deep collaborative analysis for each city conducted by multidisciplinary experts. With respect to Q2, it is clear that each existing international/national standard brings a series of benefits, but there could be inherent disadvantages for adopting fixed solutions, because some could not fit to the local conditions or could induce bigger specific risks-e.g. cyber-threats, therefore, this subject must be also treated competently, similarly to Q1 case. Finally, to answer Q3, it is certain that cities cannot wait too much, so as not to confront themselves with economic, social, territorial, political, or environmental bottlenecks, jams, disasters or even points of no return, conclusively, they must address and start implementing Q1 and Q2 as soon as possible.
As for the community's position regarding smart city solutions, it is clear that it depends to a high extent on the local status-quo, e.g. the overall positive survey results mentioned in the previous chapter may be correlated with the fact that Timis is one of the most advanced counties in Romania, according to all official KPI. The situation can be though improved anytime and anywhere by informing, training, and involving the residents in the actual steps of their conceptualization, implementation, and operation.
A way to handle this complexity of factors is to own or hire a research-developmentinnovation center-RDI center, for developing and deploying the necessary specific smart solutions. Moreover, in terms of joining additional forces, hiring businesses and other stakeholders in various forms of collaboration such as public-private partnerships is essential (Osborne Clarke, 2015, Emerson et al., 2012). Private companies can faster stimulate and guide the transformation of cities into smarter cities, according to their own economic interests, which must be in a win-win relation with the interests of the local community.
All in all, the envisioned solutions should consider all fore-mentioned aspects and the human-machine dependability balance, so that people and societies develop harmoniously, healthily, and without risks.

Limitations and Suggestions for Future Research
Each of the four results subchapters of this paper indicates existing gaps and further research actions which are implied or proposed by the authors.

Conclusions
Smart systems to be deployed in the cities of twenty-first century are very complex, therefore, they must be characterized by inspired visions, competent strategies, high competencies, and wise developments. Smart city systems should be protective and resilient, and, furthermore, they should be designed in such a way that they learn and adapt their behavior to current realities and necessities.
Some smart city international standards at different abstraction levels are under development or already exist, but unfortunately certain important aspects are not covered at present. More questions rise, with respect to what set of minimal smart components should be installed in the cities?, how much standardized the solutions should be?, and how much time cities should wait for their full standardization and technological development?, given the high pressure which is manifesting in multiple spheres of day by day life.
A good way to overcome some of the actual issues, take the upcoming opportunities, learn, and evolve faster, is to interconnect the cities themselves by smart common platforms. However, this solution implies that their internal systems are compatible, therefore, developed with this purpose in mind from the beginning.
Nine major dimensions of smart city systems identified by this paper make the selection of development choices very difficult for governors, and, if one adds the complexity of local status-quos, as well as the necessity of obtaining local community's and direct stakeholders' acceptance, the challenges are then acknowledged at their full amplitude.
Therefore, the collaboration with RDI centers and private partners seems to be a necessity, as solutions must fit to all beforementioned factors, which are ever changing in the globalizing world. As such, all requirements of smart city systems must be very well identified, designed, implemented, deployed, maintained, adapted, and improved permanently.
All in all, by comprehending all aspects mentioned before, a macrosystemic top-down approach for developing smart cities, whilst considering the local specificities, would clearly bring benefits to those cities on the long run.