A Proposal for an Integrated Smart Home Service Platform

Robert Wehlitz

, Theo Zschörnig

and Bogdan Franczyk

2,3

Institute for Applied Informatics (InfAI), Goerdelerring 9, 04109 Leipzig, Germany

Information Systems Institute, Leipzig University, Grimmaische Str. 12, 04109 Leipzig, Germany

Business Informatics Institute, Wrocław University of Economics, ul. Komandorska 118-120, 53-345 Wrocław, Poland

Keywords: Internet of Things, Smart Home, Smart Services.

Abstract: The growing popularity of Internet-connected smart devices in consumers’ homes has led to a steady increase

in the number of smart homes worldwide. In order to provide meaningful added value for consumers and

businesses alike, smart devices need to be accompanied by services and applications which excel conventional

devices’ usages. However, integrated solutions to support the development and operation of smart home

services for, and across, different devices and application scenarios are still missing in research as well as in

industry. In this paper, we present the motivation, requirements and an initial concept for a fully integrated

smart home service platform (SHSP), which may serve as a basis for further discussion.

1 INTRODUCTION

The number of Internet-connected smart devices

equipped with sensors, actuators or tags is rapidly

increasing (IHS Markit, 2018). It is estimated that

worldwide about 136.4 million households will have

smart devices in 2019 (Statista, 2018) and can

therefore be referred to as smart homes. According to

Aldrich (2003), “A ‘smart home’ can be defined as a

residence equipped with computing and information

technology which anticipates and responds to the

needs of the occupants […]”. The latter part of the

citation refers to context awareness as a system

property for enabling adaptive smart homes without

explicit user interaction (Han and Lim, 2010). To

accomplish this, the basic sensing and actuation

capabilities of different smart devices have to be

utilized in a way that allows for the creation of more

intelligent smart home services and applications

(Eom et al., 2013).

The smart home market in general offers great

potential for both, businesses and customers.

Traditional energy companies, for example, have the

possibility to become smart home service providers,

and thus open new sources of income. Moreover, the

satisfaction and loyalty of energy customers can be

improved by offering smart home services which, in

addition to the provision of energy, yield added value.

Therefore, customers may benefit in many different

application scenarios, such as home energy

management, entertainment, healthcare, security and

comfort (Alaa et al., 2017).

However, an integrated platform that supports the

development and operation of context-aware smart

home services for, and across, different devices and

application scenarios is still missing (Stojkoska and

Trivodaliev, 2017).

Against this background, the main contribution of

this paper is an initial concept for a fully integrated

smart home service platform (SHSP) that aims to fill

this gap. Our objective is to share and discuss first

ideas with the scientific community and to present an

initial overview of the platform architecture.

The remainder of this paper is organised as

follows. First, we describe the motivation for our

research (Sect. 2). We then discuss related work in

this field (Sect. 3) and outline the requirements for

our proposed platform solution (Sect. 4). Afterwards,

we introduce our initial SHSP concept and describe

its main components as well as their interactions

(Sect. 5). Finally, the paper concludes with a short

summary and outlook (Sect. 6).

2 MOTIVATION

In a nutshell, the Internet of Things (IoT) consists of

heterogeneous smart devices that are capable of

communicating with one another over the Internet

(Han et al., 2016). However, smart devices, in

630

Wehlitz, R., Zschörnig, T. and Franczyk, B.

A Proposal for an Integrated Smart Home Service Platform.

DOI: 10.5220/0007751006300636

In Proceedings of the 21st International Conference on Enterprise Information Systems (ICEIS 2019), pages 630-636

ISBN: 978-989-758-372-8

comparison to conventional devices, do not offer

added value to customers if the possibilities of their

sensing and actuation capabilities are not fully

exploited. Therefore, more intelligent services which

are built on the aforementioned capabilities are

required in order to create customer value around

products like smart meters, smart lights, smart locks,

etc. (Eom et al., 2013).

With regard to the smart home market, there are

many devices available which produce data about

their surroundings (e.g. brightness, motion, humidity,

etc.) and some of them also enable the remote or

automated control of actuators, e.g. smart thermostats

for changing the room temperature (Bing et al., 2011).

However, most of these smart devices are

proprietary solutions that have no or just little

interoperability with other products (Gajewski et al.,

2017). In consequence, they can usually only be used

with the web or mobile applications of the respective

vendor. This leads to lock-in effects and complicates

the development of smart home services (Eom et al.,

2013). Furthermore, current smart home solutions

only allow for home automation based on predefined

rules, requiring users to configure smart home

services manually with regard to their individual

needs and preferences (Xu et al., 2016). This

contradicts with the statement in Sect. 1 that a smart

home should be able to learn from the behaviour of

its inhabitants in order to respond to their needs.

In this context, we see an integrated SHSP as an

important building block to provide the infrastructure

and software tools necessary to develop, offer and run

context-aware services for, and across, different

smart devices and application scenarios in adaptive

smart home environments.

3 RELATED WORK

In scientific literature, publications dealing with

service platforms dedicated to smart home

environments are rare and they often focus primarily

on security aspects (e.g. Gajewski et al., 2017;

Fernandes et al., 2016) or specific application

scenarios, such as energy management. Han and Lim

(2010), for example, make a proposal for a smart

home energy management system based on ZigBee

networks. It relies on a sensing infrastructure for

receiving and storing sensor data, an information

extractor for converting sensor data into context

information (e.g. the current location based on

geodata) and a service extract engine for context

reasoning. However, their concept addresses the

context-based invocation of smart home services

rather than real context-aware services. Al-Ali et al.,

(2017) also propose a smart home energy

management system and combine IoT technologies,

big data analytics as well as business intelligence

tools in order to reduce energy consumption in

households. Following the business intelligence

approach, they primarily focus on data visualisation

and reporting functions, also not addressing

context-aware smart home services.

Additionally, there are also publications that deal

with concepts for application-independent smart

home solutions. Gu et al., (2011), for example,

present the design of a cloud-based smart home

platform that uses an intelligent gateway for protocol

conversion and the abstraction of smart home

networks, thus enabling inter-device communication.

Ye and Huang (2011) propose a general framework

for cloud-based smart homes and describe possible

application scenarios. Bing et al., (2011) present a

smart home system architecture that consists of a

sensing and actuator layer, a gateway-based network

layer and an application layer running on a remote

management platform. Eom et al., (2013) describe a

framework for an integrated platform which includes

middleware functionality as well as data storage and

ingestion capabilities in order to provide smart home

services with real-time data processing on cloud

platforms. Li et al. (2013) introduce a smart home

service framework based on event matching. It

consists of an IoT, data, event and service layer. The

IoT layer uses a gateway to receive smart device data.

The data layer stores and processes real-time sensor

data. The event layer identifies and matches events

based on data received from the data layer and

invokes services accordingly.

In summary, the aforementioned approaches are

primarily cloud-based, rely on gateway topologies

and provide some sort of middleware functionality.

They often allow for data processing as well as

visualisation, the remote control of smart devices or

simple home automation based on explicit user

configuration. However, almost all of them lack

advanced analytics capabilities and an integrated

concept for the development and operation of real

context-aware smart home services.

4 PLATFORM REQUIREMENTS

In this section, we describe the requirements for an

integrated SHSP. These are partly derived from the

previous work on general requirements for IoT

platforms by Wehlitz et al., (2017) and extended

using the results of workshops and expert interviews

A Proposal for an Integrated Smart Home Service Platform

631

with stakeholders of the energy industry. The

requirements are divided into functional and

non-functional requirements. The functional ones are:

 Interoperability: Different hardware, operating

systems, service interfaces, message protocols

and data formats have led to a far-reaching

heterogeneity in the IoT and smart home domain

(Issarny et al., 2011). Hence, the SHSP is to

provide appropriate means to simplify device

integration and inter-device communication in

order to overcome this heterogeneity (Wehlitz et

al., 2017).

 Event Detection and Insight Generation: The

full potential of smart device data is not yet

exploited. Therefore, the SHSP is to provide the

required infrastructure and powerful tools to

identify events in data streams as well as to

process and analyse real-time data. The main

objective is to automatically generate insights that

benefits users in many application scenarios.

 Context-aware Device Control: Current smart

home solutions often only allow for the rule-based

scheduling of device control tasks. In contrast, the

SHSP is to orchestrate sensing and actuation

services as well as advanced analytics services for

an automated and context-aware control of smart

devices (Wehlitz et al., 2017).

 Cloud- and Fog-based Deployment:

Cloud-based smart home systems predominate the

market. In the context of the SHSP, smart home

services are to be deployed and run either in the

cloud or on devices in the proximity of the users.

Which option is selected depends on the

respective application scenario, external

regulations, user preferences or locally available

computing resources.

 Service Publication: The SHSP is to offer

various capabilities to publish user-generated

artefacts. Both, customers and businesses, are to

be enabled to share smart home services with

other platform users. This aims to reach a critical

mass of platform users faster, foster innovation

based on previous work and to allow for new

business models in the smart home domain

(Wehlitz et al., 2017).

The non-functional requirements are:

 Usability: The SHSP is to provide suitable user

interfaces in order to increase the acceptance of

developers and users. In general, it is to be

designed for making the development of smart

home services more accessible to non-technical

audiences.

 Scalability: The number of platform users and

integrated smart devices as well as the amount of

sensor data to be processed may vary from small

to very large, depending on the application

scenario. Hence, the architecture of the SHSP is

to be flexible and scalable in order to meet the

performance requirements of many different

smart home applications (Wehlitz et al., 2017).

 Privacy and Security: Smart homes produce

massive amounts of personal data. These as well

as the control functions of smart devices need to

be protected against unauthorized access and

misuse. In this context, the SHSP is to enable data

owners to define fine-grained security

restrictions. This implies that users can decide

which data is accessible by services, applications

or other platform users (Wehlitz et al., 2017).

5 PROPOSAL

In this section, we introduce our initial concept for an

integrated SHSP that was designed to meet the

requirements we defined in Sect. 4. A schematic

overview of it is given in Figure 1. The main

components of the concept are: hybrid platform

architecture, middleware, streaming analytics and

machine learning, context-aware business processes,

marketplace, and identity and access management.

Following, we describe them and their interactions.

5.1 Hybrid Platform Architecture

The SHSP aims to deploy and run smart home

services both, centrally in a data centre and

decentrally in the proximity to the users.

The centralised approach addresses the cloud

computing paradigm, in which local data is

transferred via the Internet to a cloud platform for

further processing by cloud services. Cloud

computing has established itself as a disruptive

concept for the scalable provision of IT resources

(storage, network, computing power, etc.) over the

Internet without having to operate and maintain them

locally. However, this concept requires a high degree

of robustness and availability of cloud platforms, a

reliable Internet connection and a high level of user

trust in cloud providers in terms of privacy and

security.

The decentralised approach addresses the fog

computing paradigm which extends cloud computing

by making computing resources (fog nodes) available

at different hierarchical levels between the cloud and

end devices. The main idea of fog computing is to

ICEIS 2019 - 21st International Conference on Enterprise Information Systems

632

Figure 1: Overview of the SHSP concept.

move data processing closer to the “edge of the

network”. According to (Byers, 2017), this may result

in significant benefits such as higher degree of data

privacy and security, lower latency and network

traffic, as well as greater fault-tolerance, which are

important requirements in certain application

scenarios, e.g. healthcare or home security. However,

smart devices often prove to have limited computing

resources, making them unsuitable for more complex

computing tasks. Further challenges in the field of fog

computing include, among others, the design of

suitable interfaces for the provision of services on

distributed IT resources as well as the deployment

and synchronisation of cloud- and fog-based

operations (Shanhe et al., 2015).

Against this background, our SHSP concept

flexibly combines the advantages of both, cloud

computing and fog computing, within a hybrid

platform architecture.

5.2 Middleware

The middleware manages the registration and

administration of smart devices as well as the

communication between the fog and the cloud level

of the SHSP architecture. In the context of smart

home, smart devices are often connected to local

smart home hubs or gateways which provide

higher-level interfaces for accessing sensing and

actuation services (Stojkoska and Trivodaliev, 2017).

The middleware acts as the central point of

integration for smart home hubs and uses a data

abstraction model to handle heterogenous data

streams. It translates incoming messages (e.g. sensor

data) into a unified platform-internal message format

and makes the data available to the streaming

analytics and machine learning component. The

middleware also retranslates outgoing messages (e.g.

control commands to smart devices) into

device-specific requests that are handled by smart

home hubs.

In addition to ensuring the aforementioned

syntactic interoperability between smart devices, the

middleware also supports their semantic integration.

This includes the enrichment of smart device data

with metadata which allows to automatically deduce

their meaning. In order to achieve this, suitable

description languages (e.g. RDF), efficient storage

solutions (e.g. Triplestore) and query languages (e.g.

SPARQL) are used. Due to the linking of semantic

aspects, ontologies as representations of knowledge

are created and context information based on them

can be inferenced. Hereby, the management,

processing and analysis of data, as well as the

A Proposal for an Integrated Smart Home Service Platform

633

context-aware control of smart devices are to be

supported.

5.3 Streaming Analytics and Machine

Learning

In smart home application scenarios, data usually

arrives in data streams and is characterized as time

series data, thus creating the need for real-time

processing (Pawar and Attar, 2016). While the batch

processing of historical data combined with the

application of prediction models on this data is

already established, real-time processing of smart

device data, while enabling the detection of causable

relationships between the devices themselves and

their surroundings, is not. Nonetheless, this is the

basis for the autonomous adaption to new situations

(Stolpe et al., 2016).

Regarding the SHSP, the application of streaming

analytics and machine learning technologies is the

basis for developing context-aware smart home

services. Hence, it is to provide the infrastructure and

tools for sophisticated data manipulation, efficient

persistence, event processing functionalities and the

process of insight generation. The SHSP aims to

enable predictive and prescriptive analytics based on

machine learning so that the value of information

gained from historical or real-time data is increased

and thus also the benefits for smart home users.

Furthermore, tools for the automated configuration of

smart home services with regard to individual

customer needs and preferences are to be provided.

Considering the technical implementation, we

propose an approach based on the Kappa architectural

concept at the cloud level extended by the fog

computing paradigm, thus offering data processing

close to the source of data. Streaming analytics and

machine learning capabilities are encapsulated by

so-called analytics operators. These represent

microservices which, as single processing steps, can

be combined to analytics pipelines in order to solve

more complex analytics tasks. Analytics pipelines are

science via a graphical composition tool and their

results can be accessed by context-aware business

processes. In order to achieve this, a key aspect of the

streaming analytics and machine learning component

is the orchestration of analytics operators in the

overall platform architecture. Since the processing of

smart home data may be subject to different external

regulations and conditions (Byers 2017; Mouradian et

al., 2018), it is necessary to only allow for analytics

operator deployment at the fog or cloud level if

predefined deployment rules are met. These rules, for

example, could be based on hardware constraints at

the fog level and compliance regulations for the cloud

level.

5.4 Context-aware Business Processes

Within the SHSP concept, the service logic of smart

home services is defined by the notion of business

processes. Business processes primarily include

tasks, events and decisions by which sensing and

actuation services as well as services for accessing

analytics pipelines are being orchestrated. The SHSP

is to enable the modelling, implementation, execution

and monitoring of context-aware business processes.

Based on the results provided by analytics pipelines,

the control flow of business processes can be affected

at runtime, thus allowing them to adapt to new

situations. The modelling of business processes

provides the basis for the composition, reuse and

sharing of smart home services as well as their

instantiation for different smart home environments

and application scenarios. Additionally, there are

numerous tools by which they can be modelled

graphically making the development of them more

accessible to non-technical audiences. Regarding the

deployment of context-aware business processes, the

SHSP allows for both, running them at cloud as well

as at fog level. Moreover, an application

programming interface is provided which supports

the development of third-party applications, e.g.

dashboards or mobile applications.

5.5 Marketplace

The major feature of the marketplace is to allow

platform users (customers and businesses) of the

SHSP to easily share and subscribe to user-generated

artefacts which include data, analytics operators and

pipelines, as well as business process models. In this

regard, users may decide if their artefacts are

published directly on the marketplace or only be

shared with single users or user groups. This is to

render the platform attractive to developers and to

foster innovative smart home services as a result of

reusing and adapting already existing components.

Furthermore, the marketplace has the potential to

encourage new innovative business models in the

smart home domain.

5.6 Identity and Access Management

The identity and access management system are an

important building block for registering smart devices

and managing platform users and access rights across

ICEIS 2019 - 21st International Conference on Enterprise Information Systems

634

the different levels of the platform architecture. They

govern the authentication and authorization of users

and enable fine-grained access control for smart

devices, data, analytics as well as context-aware

business processes. In this context, the identity and

access management system can also be seen as the

basis for publishing smart home services on the

marketplace or sharing them with other platform

users.

6 CONCLUSIONS

In this paper, the motivation, requirements and initial

concept for an integrated SHSP, which offers tools to

develop and run context-aware smart home services

are presented. The concept is based on a hybrid

platform architecture extending cloud computing

with the fog computing paradigm. A middleware

ensures the syntactic and semantic interoperability of

smart devices. Streaming analytics and machine

learning capabilities in the form of analytics operators

and pipelines are used to provide meaningful insights

into smart device data. The service logic of smart

home services is defined by the means of

context-aware business processes, whereby the SHSP

provides tools for their modelling, implementation,

execution and monitoring. A marketplace offers the

possibility to share user-generated artefacts. The

definition and application of fine-grained access

control policies to preserve privacy and security is

enabled by the identity and access management

system.

In the near future, the individual areas of the

SHSP need to be investigated deeper in terms of their

detailed design and technical implementation as well

as focusing on the interaction between them. It is

essential that all subsystems of the platform are

designed in such a manner that it is easy to use also

for non-technical audiences. Finally, the proposed

hybrid platform architecture approach needs to be

investigated further and, if feasible, evaluated in

terms of performance compared to similar

approaches.

ACKNOWLEDGEMENTS

The work presented in this paper is partly funded by

the European Regional Development Fund (ERDF)

and the Free State of Saxony (Sächsische

AufbauBank – SAB).

REFERENCES

Alaa M., Zaidan A.A., Zaidan B.B., Talal M., Kiah M.L.M.,

(2017). A review of smart home applications based on

Internet of Things. In Journal of Network and

Computer Applications, 97, 48-65.

Al-Ali, A.R., Zualkernan, I., Rashid, M., Gupta, R.,

Alikarar, M., (2017). A smart home energy

management system using IoT and big data analytics

approach. In IEEE Transactions on Consumer

Electronics. 63, 426-434.

Aldrich, F.K., (2003). Smart Homes: Past, Present and

Future. In Inside the Smart Home. Springer, London.

Bing, K., Fu, L., Zhuo, Y., Yanlei, L., (2011). Design of an

Internet of Things-based smart home system. In

International Conference on Intelligent Control and

Information Processing (ICICIP), 921-924.

Byers C. C., (2017). Architectural Imperatives for Fog

Computing: Use Cases, Requirements, and

Architectural Techniques for Fog-Enabled IoT

Networks. In IEEE Communications Magazine, 55 (8).

Eom, B., Lee, C., Yoon, C., Lee, H., Ryu, W., (2013). A

Platform as a Service for Smart Home. In International

Journal of Future Computer and Communication. 253-

257.

Fernandes, E., Jung, J., Prakash, A., (2016). Security

Analysis of Emerging Smart Home Applications. In

IEEE Symposium on Security and Privacy (SP), 636-

654.

Gajewski, M., Batalla, J., Mastorakis, G., Mavromoustakis,

C., (2017). A distributed IDS architecture model for

Smart Home systems. In Cluster Computing.

Gu, H., Diao, Y., Liu, W., Zhang, X., (2011). The design of

smart home platform based on Cloud Computing. In

International Conference on Electronic and

Mechanical Engineering and Information Technology

(EMEIT), 3919-3922.

Han, S.N., Khan, I., Lee, G.M., Crespi, N., Glitho, R.H.,

(2016): Service composition for IP smart object using

realtime Web protocols: Concept and research

challenges. In Comput. Stand. Interfaces, 43, 79-90.

Han, D.-M., Lim, J., (2010). Design and Implementation of

Smart Home Energy Management Systems based on

ZigBee. In IEEE Transactions on Consumer

Electronics, 56, 1417-1425.

IHS Markit, (2018). IoT trend watch 2018, White paper.

Issarny, V., Georgantas, N., Hachem, S., Zarras, A.,

Vassiliadist, P., Autili, M., Gerosa, M.A., Ben Hamida,

A., (2011). Service-oriented middleware for the Future

Internet: state of the art and research directions. In

Journal of Internet Services and Applications, 2, 23-45.

Li, Q., Jin, Y., He, T., Hanchuan, Xu, (2013). Smart home

services based on event matching. In International

Conference on Fuzzy Systems and Knowledge

Discovery (SKD), 762-766.

Mouradian, C., Naboulsi, D., Yangui, S., Glitho, R. H.,

Morrow, M. J., & Polakos, P. A. (2018). A

Comprehensive Survey on Fog Computing: State-of-

the-Art and Research Challenges. In IEEE

Communications Surveys & Tutorials, 20, 416-464.

A Proposal for an Integrated Smart Home Service Platform

635

Pawar, K., & Attar, V., (2016). A survey on Data Analytic

Platforms for Internet of Things. In: CAST-2016, 605-

610.

Shanhe Y., Cheng L., Qun L., (2015). A Survey of Fog

Computing: Concepts, Applications and Issues. In

Proceedings of the Workshop on Mobile Big Data

(Mobidata).

Statista, (2018). Smart Home Report 2019, Statista Digital

Market Outlook – Market Report.

Stojkoska B.L.R., Trivodaliev K.V., (2017). A review of

Internet of Things for smart home: Challenges and

solutions. In Journal of Cleaner Production, 140 (3),

1454-1464.

Stolpe, M., (2016). The Internet of Things: Opportunities

and Challenges for Distributed Data Analysis. In ACM

SIGKDD Explorations Newsletter, 18(1), 15-34.

Wehlitz, R., Häberlein, D., Zschörnig, T., Franczyk, B.,

(2017). A Smart Energy Platform for the Internet of

Things - Motivation, Challenges, and Solution

Proposal. In Business Information Systems, 271-282.

Xu, K., Wang, X., Wei, W., Song, H., Mao, B., (2016).

Toward software defined smart home. In IEEE

Communications Magazine, 54, 116-122.

Ye, X., Huang, J., (2011). A framework for Cloud-based

Smart Home. International Conference on Computer

Science and Network Technology (ICCSNT), 894-897.

ICEIS 2019 - 21st International Conference on Enterprise Information Systems

636