[Paper Review] Microservice-based IoT for Smart Buildings

Dilshat Salikhov, Kevin Khanda+3 more|arXiv (Cornell University)|Oct 29, 2016

Context-Aware Activity Recognition Systems14 references4 citations

TL;DR

This paper proposes a microservice-based IoT platform for smart buildings using the Jolie programming language and a distributed sensor network to enable scalable, flexible, and interoperable integration of diverse devices. The prototype, deployed in a real academic building, demonstrates seamless coordination of sensors (temperature, motion, light, door, camera) and actuators via modular microservices, achieving real-time environmental monitoring and adaptive control with potential for energy savings and improved occupant comfort.

ABSTRACT

A large percentage of buildings in domestic or special-purpose is expected to become increasingly "smarter" in the future, due to the immense benefits in terms of energy saving, safety, flexibility, and comfort, that relevant new technologies offer. As concerns hardware, software, or platform level, however, no clearly dominant standards currently exist. Such standards, would ideally, fulfill a number of important desiderata, which are to be touched upon in this paper. Here, we will present a prototype platform for supporting multiple concurrent applications for smart buildings, which is utilizing an advanced sensor network as well as a distributed microservices architecture, centrally featuring the Jolie programming language. The architecture and benefits of our system are discussed, as well as a prototype containing a number of nodes and a user interface, deployed in a real-world academic building environment. Our results illustrate the promising nature of our approach, as well as open avenues for future work towards its wider and larger scale applicability.

Motivation & Objective

• To address the lack of standardized, open, and extensible platforms for smart building automation that support diverse, concurrent applications.
• To develop a flexible, scalable, and interoperable infrastructure for integrating heterogeneous IoT devices (sensors and actuators) in real building environments.
• To evaluate the feasibility and benefits of using the Jolie programming language for implementing distributed microservices in building automation systems.
• To enable real-time data collection and adaptive control of environmental parameters (temperature, lighting, occupancy) to improve energy efficiency and occupant comfort.
• To lay the foundation for a unified, extensible platform supporting Human-Building Interaction (HBI) and future AI-driven optimization.

Proposed method

• Design and implement a distributed microservices architecture using the Jolie programming language to manage communication and coordination between IoT devices.
• Deploy a heterogeneous sensor network including SensorTags (for temperature, motion, light, humidity), Raspberry Pi as a central node, and a camera for occupancy detection.
• Integrate multiple communication protocols (ZigBee, BLE, Z-Wave) via custom Jolie modules to enable device interoperability.
• Use a centralized control system to collect sensor data, process it in real time, and trigger actuation (e.g., lighting, power sockets) based on occupancy and preferences.
• Implement a Telegram-based feedback bot to gather real-time comfort preferences from users, enhancing personalization.
• Collect and analyze data over a one-month period to train a predictive model for optimizing environmental settings.

Experimental results

Research questions

• RQ1 Can a microservices-based architecture using Jolie effectively integrate heterogeneous IoT devices (sensors and actuators) in a real smart building environment?
• RQ2 How can a unified platform support concurrent, independent applications (e.g., energy management, occupancy tracking, comfort feedback) while sharing the same infrastructure?
• RQ3 To what extent can real-time sensor data and user feedback be used to predict and optimize environmental conditions (temperature, lighting, humidity) for improved comfort and energy savings?
• RQ4 What are the practical challenges in deploying a scalable, secure, and maintainable IoT platform in an academic building using open-source tools?
• RQ5 How can the platform evolve to support self-configuration and adaptive behavior in response to changing environmental and human needs?

Key findings

• The Jolie-based microservices architecture successfully enabled real-time, interoperable communication between diverse IoT devices (ZigBee, BLE, Z-Wave) using a single, unified programming model.
• The prototype demonstrated reliable data collection from multiple sensors (temperature, motion, light, door, camera) and effective control of lighting and power sockets via relays.
• A Telegram-based feedback system successfully gathered real-time comfort preferences from students and staff, enabling data-driven personalization of environmental settings.
• The system collected a one-month dataset on temperature, humidity, lighting, and occupancy, forming the basis for future predictive modeling of optimal indoor conditions.
• The platform proved scalable and extensible, with clear pathways for integrating new devices and protocols, and for deploying AI-driven optimization in future iterations.
• The deployment in a real academic building validated the feasibility of the approach for real-world smart building applications, with strong potential for energy savings and enhanced occupant experience.