An internet of things service roadmap

Athman Bouguettaya, Quan Z. Sheng, Boualem Benatallah, Azadeh Ghari Neiat, Sajib Mistry, Aditya Ghose, Surya Nepal, Lina Yao

Introduction

The Internet of Things (IoT) has successfully transitioned from a theoretical concept to a global reality, connecting billions of sensors and actuators across various domains. However, the field is currently at a "technology" stage, characterized by fragmented systems and a focus on the underlying "plumbing" — the mechanics of device connectivity and data transmission. While we can move data, we often struggle to make it useful. This paper argues that the next revolution in IoT requires a shift to a "service" stage, much like the evolution of the Internet into the World Wide Web. By adopting a service-oriented paradigm, researchers can move beyond raw data collection to create a scalable, intelligent, and human-centric ecosystem of actionable knowledge.

The Service Paradigm: An Analogy-Based Framework

The core of this roadmap is the transition from viewing IoT as a collection of "things" to viewing it as a suite of "services." In this model, hardware functions are treated as abstract, composable services. A service description goes beyond raw data, including both functional attributes (what the device does) and non-functional attributes, such as Quality of Service (QoS). This abstraction allows for "plug-and-play" interoperability between heterogeneous devices.

The authors frame this transition through a conceptual analogy comparing the evolution of the Web to the future of IoT:

  • The Internet vs. The Web: The Internet provided the foundational communication technology, but the Web transformed it into a meaningful platform for users through democratization, commoditization, and digitization.
  • IoT vs. IoT Services: Similarly, while the IoT provides the hardware foundation, "IoT Services" will provide the necessary framework through three new pillars: Smartization (adding intelligence), Augmentation (enhancing human experience), and Contextualization (providing situational awareness).
The conceptual evolution from the Web's pillars to the proposed IoT service pillars.

Ten Critical Research Frontiers

To realize the vision of a "Web of Things," the authors identify ten major research challenges that the community must address:

  1. Open Actuation: Most current research focuses on sensing (observing the world). The future lies in actuation — using the Internet to trigger physical state changes. "Open actuation" allows operators to identify and utilize nearby devices on the fly to achieve specific goals.
  2. Servitization: This involves wrapping physical products in service-oriented models, allowing them to be managed and composed more effectively.
  3. Discovery and Ranking: With billions of dynamic, moving devices, the IoT needs a "natural order ranking" system — similar to Google’s PageRank — to help users find the most relevant services.
  4. Security, Privacy, and Trust: As devices become more integrated into human life (e.g., brain wearables), we need lightweight, quantum-resistant security protocols that can run on resource-constrained hardware.
  5. Crowdsourcing: Managing trust and data quality in decentralized, transient networks of devices is a significant hurdle.
  6. Experiential IoT: Moving from "data analytics" to "experience analytics." This involves capturing the four dimensions of human experience — time, space, actors, and things — to allow experiences to be digitally represented, transferred, or revisited.
  7. Requirements-driven Design: Developing methods to map complex, high-level human needs to specific configurations of sensors and actuators.
  8. Complex Compositions: Orchestrating massive, heterogeneous groups of interchangeable sensors to perform complex tasks.
  9. Large-Scale Facilities: There is an urgent need for nation-wide, real-world testbeds, such as the IoT-Lab in France, to generate the data necessary for empirical research.
  10. Data Analytical Services: Solving the "tangled relationships" within multi-dimensional IoT data to extract meaningful insights.

Challenges and Future Outlook

Despite the clear vision, several formidable hurdles remain. Scalability is a primary concern; current discovery and security solutions do not yet scale to the billions of devices expected in a global ecosystem. Heterogeneity also poses a challenge, as there is currently no single ontological standard for describing IoT services, making seamless interoperability difficult.

Furthermore, many IoT devices are resource-constrained, lacking the power to run modern security protocols, which necessitates the development of specialized "lightweight" alternatives. Finally, as IoT data increasingly informs AI, ensuring these systems are free from algorithmic bias is a critical, yet unsolved, challenge.

Conclusion

The vision of a "Web of Things" is clear, but its realization requires significant progress in service engineering. By shifting the focus from the "management of things" to the "provision of services," the IoT can finally fulfill its potential as a provider of actionable knowledge and enhanced human experiences.

Summary updated

BibTeX
@article{bouguettaya2021internet,
  author = {Athman Bouguettaya and Quan Z. Sheng and Boualem Benatallah and Azadeh Ghari Neiat and Sajib Mistry and Aditya Ghose and Surya Nepal and Lina Yao},
  title = {{An internet of things service roadmap}},
  journal = {Communications of the ACM},
  year = {2021},
  doi = {10.1145/3464960},
}
Figure