The Internet of Things: The Story So Far

 

The combination of embedded technologies, wired and wireless communications and low cost sensing devices on the Internet make up the Internet of Things (IoT). With an expected 50 billion connected things by 2020, this has created huge interest. Predating the current situation in the IoT were RFID technologies for identifying real world objects, (wireless) sensor and actuator networks.

The most recent progress in IoT has resulted from industry and consumer market interest in connected sensing devices. Several products have been introduced for sports and activity monitoring, personal health monitoring (and the associated Quantified Self movement) and other consumer and retail markets. There is also a new trend of Internet and mobile software and services for monitoring and controlling personal devices and home appliances. However, these products rely on vertical and proprietary solutions that have limited interoperability with other devices and services.

Heterogeneous data and services

The IoT is evolving as a distributed, multi-vendor and multi-platform framework with heterogeneity at device, network, data and services levels. In the past few years there has been significant progress in standardising wireless communication technologies and providing efficient solutions for low power, resource-constrained IoT devices. The IETF Core standards and IPv6 over Low power Wireless Personal Area Networks (6LowPAN) and Constrained Application Protocol (CoAP) [1] are examples of these efforts. IoT data communication is becoming an integrated part of mobile communications and future generations of mobile communications and 5G networks are now being designed to support voice, text and multimedia data and also machine-to-machine communications and connection and control for IoT devices with constrained resources and intermittent data patterns. These standards and systems are increasingly being deployed in public and private sectors.

IoT research and development is now moving from infrastructure and baseline technology development, or early adoption of standalone solutions, towards the standardisation of solutions and the definition of common components and practices. However, heterogeneity at the semantic level still remains a key issue. To enable effective and automated data and service communication and interactions, data and services should be defined in common and interoperable formats. While the introduction of TCP/IP on Internet and HTTP protocols paved the way for the rapid growth of the Web and markup languages such as HTML allowed the publication of millions of pages on the Web, IoT needs its own specific or adapted and extended higher-level protocols and common formats to enable interoperability between various heterogeneous data and services. Several efforts in this area, such as the W3C Semantic Sensor Network Ontology (SSNO) [2] and HyperCat [3], have been introduced and SSNO has seen quite a few demonstrations and early adoption. However, these models need to be adapted and exploited by more products and services. There is also a need for software and development APIs to allow publishing, sharing and access based on common formats. Linking IoT data to other data on the Web and providing linked IoT data forms will also enhance the use and exploitation of the data and services. Access interfaces, query and discovery methods similar to those offered for the Internet and Web resources should be also provided in order for the IoT domain to make data and services widely accessible beyond internal IoT networks.

Actionable knowledge

The IoT is not about collecting and publishing data from the physical world but rather about providing knowledge and insights regarding objects (i.e., things), the physical environment, the human and social activities in the physical environments (as may be recorded by devices), and enabling systems to take action based on the knowledge obtained. In other words, raw IoT data is not what the IoT user wants; it is mainly about ambient intelligence and actionable knowledge enabled by real world and real time data [4]. Figure 1 shows the different waves of IoT development. As discussed above, it started with the RFID developments and is now mainly focused on physical-cyber-social data and integrated systems with various products and services and prototype models for data/service interoperability.

Figure 1. Different waves of IoT development

The IoT has become an integral part of many industry R&D units in large industries and there are growing numbers of start-ups and SMEs that focus their business models on IoT technologies. Some public sector areas have also taken a keen interest in using IoT technologies to provide better community services, healthcare, transport and environmental control and monitoring, among other applications. Several cities around the world now have plans to develop, or have already developed and exploited, IoT-based solutions in their smart city frameworks. Wearable technologies and smartphone and smart devices are driving the rapid growth and adaptation of IoT products and services in the consumer market. Industry solutions based on IoT are emerging, with some early adaptors in transportation, logistics and health. The IoT is already around us. It is not one solution or a unified technology; it involves several domains, various technologies and different coordinated and uncoordinated efforts to connect and exploit the Things’ data.

Future developments in the IoT domain are going to have a stronger focus on data and on extracting actionable knowledge and providing value-added services. This will depend on developing efficient and interoperable solutions across different platforms and various networks, and enabling semantic interoperability among various resources, data and services. Cloud-based back end services for the efficient integration, aggregation, interpretation and information extraction of multi-modal IoT data are crucial for future developments in the IoT domain. IoT data can be unreliable, incomplete and could have various qualities. The data is often time and location dependent and processing methods should be able to process and extract information from various multi-modal and real time streams and often in a (near) real time manner. This will require more adaptable and dynamic analytics solutions than the classic data mining and data analysis solutions.

Let us not forget that IoT enables the collection and dissemination of data from public and personal environments. So security, privacy and trust will always be core issues and considerations in many IoT applications and services. Industry will need service level agreements and new business models. The growing trend of social media and crowdsourcing has also enabled the concept of human sensors or ‘Citizen Sensing’ in which people use smart devices and social tools to report their observations and measurements from the physical world. Discovering, integrating and interpreting these various multi-modal physical-cyber-social streams, providing timely and sufficiently accurate and reliable insights and actionable knowledge from the data are among the key challenges. IoT solutions should consider resource and network characteristics and limitations (e.g., energy efficiency, latency), quality issues (e.g., quality of information and quality of services), and should provide global, scalable solutions that go beyond the vertical networks and offer reliable and dependable services and applications for both consumer and industry markets. Future IoT technologies need to be able to translate the deluge of dynamic and heterogeneous data from the large number of connected devices into situational awareness, actionable information and better decisions leading to improved productivity and better quality of life.

Acknowledgement

The authors are funded in part by the EU FP7 CityPulse project (Contract Number: CNECT-ICT-609035).

References

[1] Bormann, C.; Castellani, AP.; Shelby, Z., "CoAP: An Application Protocol for Billions of Tiny Internet Nodes," Internet Computing, IEEE, vol.16, no.2, pp.62-67, March-April 2012

[2] Compton; M., et al., "The SSN ontology of the W3C semantic sensor network incubator group", Journal of Web Semantics, 17, Dec. 2012

[3] HyperCat, available at: http://wiki.1248.io/doku.php?id=hypercat

[4] Barnaghi, P.; Sheth, A; Henson, C., "From Data to Actionable Knowledge: Big Data Challenges in the Web of Things," Intelligent Systems, IEEE, vol.28, no.6, pp.6-11, Nov/Dec. 2013

 

---

 

Payam Barnaghi is a Lecturer (Assistant Professor) at the Institute for Communication Systems at the University of Surrey. He is technical coordinator of the EU FP7 CityPulse project (http://ict-citypulse.eu). His research interests include machine learning, Internet of Things, data analytics, semantic web, information centric networks, and information search and retrieval. Barnaghi has a PhD in Computer Science from University of Malaya. He is a senior member of IEEE.
Contact him at: p.barnaghi@surrey.ac.uk;
http://personal.ee.surrey.ac.uk/Personal/P.Barnaghi

 

Amit Sheth is the LexisNexis Ohio Eminent Scholar and director of Kno.e.sis at Wright State University. His research interests include Web 3.0 (including the Semantic Web), semantics-empowered social Web, sensor Web, the Web of things, mobile computing, and cloud computing. Sheth has a PhD in computer and information science from Ohio State University. He is a fellow of IEEE.
Contact him at amit@knoesis.org; http://knoesis.org/amit

 

Comments