Around the world, growing urbanization has led to crowding in cities. Consequently, urban infrastructure has fallen short and citizen service quality has been poor. Citizens have disengaged from government because citizen services have often been inaccessible; and when accessible, merely transactional; and, at best, impersonal. TCS Research, on the contrary, has reimagined the urban center as a Smart City undergirded by personalization, with lifelong citizen enga gement as its touchstone, and citizen-centrism as its key measure. It has conceptualized the building of the smart city via an artificial intelligence (AI)–based approach, followed by implementation of an integrated, technology. The success of this blueprint is in evidence in the Dubai Design District, Dubai and at the SHINESeniors assistedliving housing project, Singapore.
Around the world, urbanization is surging. Urban life offers a host of benefits to citizens and promises them what they could never hope for in rural areas. Yet the rapid rise in an already high urban population density poses enormous challenges for civic administrations that try to create an all-inclusive, efficient, collaborative, and intelligent engagement for the citizen with urban services and infrastructure. Over recent decades, one has been witnessing these challenges not only in multiple sectors individually but increasingly from overlapping areas between domains.
The first wave Information and Communications Technology (ICT) enablers for governance was e-governance implementation. A multitude of government services were offered to citizens through dedicated online portals. Although that created process efficiencies, the services remained transactional. The approach failed to cater to the heterogeneity in citizens’ backgrounds and life contexts. Traditionally, the government and society had shaken hands, thanks primarily to informal intermediation. But the large scale of current operations made that inadequate. The citizen was disconnected from service discovery, while beneficiary identification by the government and its agencies had significant gaps. Many benefits planned for the citizen did not raise quality of life (QoL) indicators. Personalization, essential for the success of any human governance mechanism, was lacking.
Focusing on citizen engagement
To address these challenges, TCS has been researching the Smart City concept for many years. TCS describes the smart city as an inclusive, citizencentric urban entity that fosters a sustainable environment in which all stakeholders enjoy a high QoL, using the city’s resources optimally through technology. The Digital Citizen and Connected Social Systems Research group’s focus is to build researchdriven digital systems to assist citizens and society to improve QoL through intelligent, sustainable, and empowered interactions with the ecosystem, using data-driven approaches.
TCS’ citizen-centric Smart City concept is founded on technology-enriched sociological tenets such as real-time monitoring of assets, participative governance, citizens as cocreators, the platform-based sharing economy, citizen- and service-centric open data, and behavioral nudging for personal and public good. Together, these promise an exciting world of possibilities and can become a social disruptor in the way the urban-dweller experiences their touch points with the government and other services in a city.
We reimagine a scenario where the government comes to the citizens, rather than the citizens going to the government. This means a personalized government, where citizens are aware of what is happening in the city and what is applicable and relevant for them. Using the information already available with the government and collating it, the Smart City system recommends the right services to citizens at the right points in their lifecycle contexts, building in behavioral nudges, and orchestrates these services across private and public sectors. This revolutionizes the nature of benefit transfers— from on-and-off transactions to lifelong engagement!
The touchstone of TCS’ concept of the Smart City is lifelong citizen engagement
Smartness in a city can be looked at as an integrated cross-disciplinary approach to delivering stakeholder experience (Figure 1). Therefore, it needs to look at multiple stakeholders—the citizen, the community, the city administration, and the business. Thus, our focus spans all four stakeholders. The citizen looks for great experience with services. Communities focus on collaboration and self-help. Mayors, as representatives of civic administration, look at building a digital spine. And businesses look to transform themselves and their customers’ experience.
A three-phase AI-based approach
The touchstone of TCS’ concept of the Smart City is lifelong citizen engagement. This is ensured through a three-phase AI-based approach. To overcome limitations of human effort in providing services with accuracy and speed at scale, this approach is strongly technology-enabled. However, our AI-based approach augments the human touch, rather than replace it, and incorporates human-like lastmile delivery.
Understanding the citizen and citizen journeys
This involves sensing the citizen and understanding their context by asking “who,” “what,” “why,” and “how” questions and breaking that down to the steps a citizen takes to interact with the city’s services (the journey). Every citizen’s context has two dimensions—a real-time context, (for example, citizens’ physical locations, walking speed and other activities, health, wellness, abnormal events, city traffic conditions, and interaction with infrastructure, which cover the actual markers of the empirical state of the citizen as an entity), and a “learned” context (that includes the preferences and behavioral patterns manifested by the citizen in response to the real-time context; for example, a citizen’s mood, motivators, social interactions, and quality of participation in civic affairs). The synthesis or integration of the citizen and the journey enables us to build a citizen persona, which is the baseline for planning, designing, and delivering personalized services.
Creating a citizen-contextual model
Since the citizenry covered by a given urban center is heterogeneous, persona creation results in multiple citizen personas, which then need to be abstracted, parameterized, approximated using a hierarchy of characteristics, classified, and generalized so that a persona can be manipulated for input-output mapping. The modeling phase equips us to match the right services to citizens aspiring for them.
Matching and orchestrating the right services from government and third parties
Modeling leads to service matching and orchestration. The matching process uses AI features such as recommender systems, which provide the best-matching service options (including nudging), given the citizen’s modeled eligibility. Orchestration coordinates the various steps required to deliver the services, with real-time notifications keeping citizens informed of progress or of actions or responses expected from them. All of this is carried out not as transactional but lifetime services.
TCS uses an integrated, two-part technology for Smart City deployment: a citizen-sensing and engagement platform, and a Smart City data hub. A good example is the assisted-living project in Singapore (See Box 1 and 2). For Dubai Design District (d3), we did work on citizens and citizen journeys.
The citizen-sensing and engagement platform
To sense citizens and deliver the right services to them, ground truthing is done through actual personal or focus group interviews.
When the population surveyed is too large for face-to-face interviewing, automated sensing provides data. For instance, during a citizen’s commute in a car, details regarding traffic congestion, stream flow, and potholes are captured through gyrostabilizers and accelerometers of commuters’ smartphone.
Location, if outdoors, is sensed through GPS and, if indoors, through passive infrared (PIR) sensing, Wi-Fi triangulation, received signal strength indication, beacons, and video cameras, among other methods.
The activities of the person, a major area of our research, are detected using a variety of techniques and devices—video and eyeball tracking, sensors for opening and closing of doors or medication box lids, wearables, radio frequency (RF) sensing, and smart plugs.
Physical wellness is gauged by monitoring abnormal patterns in sensor data. For instance, no physical movement for, say, 3 hours continuously might indicate loss of consciousness. Physical frailty, which could be indicated by abnormal sleep patterns, may be inferred from extended periods spent in the bedroom. Fuzzy scores generated when a person stands in front of a Microsoft KinectTM could indicate physical instability or lack of balance.
Likewise, social wellness can be inferred by correlating current patterns with past data and ground truth validation; for example, if a person no longer steps out of the house, they may be feeling lonely.
Mental wellness can be inferred from abnormal activity patterns. For example, if a person is suddenly seen to be wandering between two rooms, it could be a sign of mild cognitive impairment.
The smart city data hub
Smart City data, much of which is generated by devices sensing the city, its assets, and citizens in real time, and by enterprise and operational systems and social media, is humongous. Thus, it is critical that the data is collected and fused, and actionable insights generated therefrom. Moreover, as a city’s capabilities grow, the risk of incoming data from individual systems being stored in silos— especially in an Internet of Things (IoT) environment—mounts rapidly.
Without a well-designed, standardized data hub, it is difficult to manage, collaborate, enhance, extend, control, and scale a city’s smart capabilities.
To store citizen data elegantly and derive insights for exploring service opportunities efficiently, TCS is building a data hub—an information platform populated with data from discrete sources, such as IoT, enterprise, and crowdsourced data. For example, a customer relationship management system can be used for a citizen’s details; Google Calendar for their doctor’s appointment; and a health management system for his/her medication plan. The data is then abstracted, contextualized, and business-enriched and, thereafter, delivered to multiple destinations— government, business, the public, and various agencies—who then derive meaning from it. All this is done in compliance with open standards.
We are applying this to a variety of use cases in domains such as wellness, transportation, and smart living.
Designing the d3 Smart City Experience
The Dubai Design District (d3) engaged TCS Research’s design team to deliver five Smart City services – smart parking, common area energy automation, street lighting automation, business partner dashboards, and management dashboards. We believed it was important to define each service from end-to-end utilizing human-centered design thinking. We adopted a “Discover, Design, and Detail” process.
In the Discover phase we first sought an understanding of the stakeholder goals, covering the Smart Dubai pillars (efficiency, seamlessness, safety, and impact) and the d3 brand promise. Then we set out to study different user segments through observation and indepth interviews to get an insight into their lives and their journeys at d3. We ascertained user perception of “smartness” and what the four pillars of smartness meant to them in the context of the services being designed. (We also conducted a comparative evaluation of similar hubs and did secondary research on smart cities and design districts to draw inspiration.)
In the Define phase we analysed all the information gathered and synthesised it to unearth insights. We identified the key personas that represent the users of d3 (Figure 3), their journeys at d3 and how the specific services intersect as they go through their journeys.
The values sought by the personas and their perceived meaning of “smart” were the two perspectives through which TCS detailed each customer journey. This helped to identify opportunities at different stages of the journey. The opportunities identified through journey maps were clustered and elaborated upon for ideation.
TCS evaluated the diverse set of ideas generated at the end of the Define stage for their relevance, uniqueness and feasibility in the Detail phase. The ideas that passed the relevance, uniqueness and feasibility criteria underwent two further stages of refinement before execution.
Scenario decomposition: Every scenario was meticulously examined and dissected into the building blocks of a service definition: actions, actors, touchpoints, affordances, channels, and artefacts.
Service definition: The decomposed scenarios were translated into a feature catalogue and divided amongst the five smart city services that could be used to provide a service definition, followed by service specification. For example, in order to understand the user expectations in terms of common services, we asked tenants to share their views on parking. Some desired a premium level of service reflecting their social status. This resulted in customization, especially in visitor management, such as personalized welcome messages, escorts/valets, and guided navigation. For other tenants, convenience was paramount, and they placed higher value on the ability to find parking spots close to their building, and hassle-free entry and exit. Some users wished to see the availability of parking spots for short-term rental communicated through online channels.
This service design methodology enabled the creation of meaningful and “smart” experiences for d3 tenants. Basing this work on design principles and using personas helped us tailor even routine services such as parking, lighting, and energy automation, to the unique culture and specific needs of each segment of d3.
Applying our research to eldercare
In addition to the d3 project, TCS Research has successfully completed citizen-sensing and pilot program for the elderly in Singapore (see Box 2).
In many countries, the change in demographics translates in an increase in the population of the elderly (those over 65 years in age), who are past the working age. Also, their healthcare costs are high because, soon after they reach that stage, they need assistance in even day-to-day living, due to various lifestyle factors.
How can Smart City technologies help the elderly to age in place, with dignity, in the safety and comfort of their own homes or an eldercare facility? How can technology non intrusively ensure their well-being?
TCS’ answer came in the form of smart-enabling many apartments in Singapore’s Housing Development Board.
Singapore’s SHINESeniors Assisted-Living Experience
TCS Research, together with Singapore Management University (SMU), deployed an assisted-living platform for the elderly. The technologies used play a key role in keeping the elderly safe at home without burdening the country’s healthcare system.
Non-intrusive sensors track the activities of the elderly and find patterns in the data collected. Deviations from the norm, based on a learned personalized threshold – such as non-movement indoors or, alternatively, being outside one’s home for extended periods, or not opening one’s medication box—are used to alert a team of community caregivers equipped with a defined protocol for response.
One core feature of the platform is the deployment of a variety of sensors for the purpose—door sensors, wall-mounted PIR sensors, sensors on medication boxes and beds, and beacons—and applying machine learning algorithms to support reactive and predictive care. Another is a workflow that notifies the concerned stakeholders so that help can reach the elderly on time.
Work is in progress on tracking gait and fall propensity, using Microsoft KinectTM devices, and wearable-based gesture recognition. Based on the activity data of the elderly, TCS and SMU have also had encouraging results finding and predicting if a person is suffering from social isolation, psychiatric conditions such as depression, or physical conditions such as sleep disorder. Fused with health data, this can yield richer insights and validations.
Privacy and security considerations
Is citizen-sensing intrusive? Not necessarily. There is a gradation of intrusiveness, from low (PIR and RF sensing), to medium (wearables sensing), to highly intrusive (video sensing, even if one preserves the privacy of the individual being sensed). Depending on the context, value to the customer, and accuracy of insights, there is a trade-off. A robust solution should support a multi-sensor fusion that uses a level of intrusiveness commensurate with the value gained.
This leads to the issue of security and privacy. Obviously, citizen data is highly sensitive and given the EU’s General Data Protection and Regulation (GDPR) mandate, for instance, we must ensure that any citizen data we store abides by GDPR or other corresponding requirements.
However, delivering personalized services necessitates access to citizen data. How does one balance personalization with security and privacy? By deploying an online avatar—a concept TCS is working on. An avatar is an agent for the citizen, which enforces an auditable privacy-compliant layer over citizen data and which any third-party service must use to access the data that is fully under the citizen’s control.
Field-Testing our research
TCS’ mechanism for validating our research in the field is through TCS Co-Innovation NetworkTM (TCS COINTM), a strong network of partners, co-creators, and early adopters that brings together the best of academia, research, tech start-ups, and venture capitalists to formalize innovation across a wide spectrum of industries.
A partnership under TCS COINTM is the iCity Lab, an SMU-TCS joint venture established in 2011 and focused on multidisciplinary R&D in intelligent, inclusive, and integrative city and societal solutions. The iCity Lab combines TCS’ industry-leading IT services expertise and culture of innovation with SMU’s globally recognized excellence in public- and private-sector research and education in business and management. Deeply embedded in TCS’ research efforts, the Lab works with multiple stakeholders, including government agencies and community partners, conducting research on the Smart City and citizens, using Singapore as a test bed. A key focus area for the Lab is the elderly, in which the Lab collaborates with TCS on jointly piloting solutions in real-world settings, validating TCS’ algorithms, and providing invaluable data for future research. The SHINESeniors initiative is one such solution.
What the future holds
Among the technologies that will play a significant role in the citizencentric Smart City are the following:
IoT. Novel forms of sensing will continue to emerge, with their own characteristics, in terms of cost, accuracy, privacy, reliability, and intrusiveness. Custom algorithms will be designed for each.
AI for public and individual good. TCS is already using machine learning to detect and predict abnormal patterns in the activities of the elderly. The next step will be to prevent abnormal events or errors.
Nudging. Nudge theory studies how sensors can influence the behavior of humans for the public good. This will soon cover other areas that can have significant impact—game theory, gamification, behavior modeling, agent-based simulations, and demand-response mechanisms
Robotics. What if the last-mile human touch were to be provided by a robot? What if the citizen were to interact with one to improve his QoL? Could a robot detect a low in citizen wellness and intervene?
Citizen experience and personalization. These technologies will explore how different citizen segments experience the Smart City and how the technologies can personalize it.
TCS has taken citizen-centrism as being the key measure of the Smart City. This will be reinforced as the Smart City proliferates, with many more challenging areas under the behavioral sciences, data sciences, and IoT opening up for research.