RIoT: A Large-Scale Real-World IoT Dataset

Stefan Petrescu, Jan S. Rellermeyer

Keywords: iot, datasets

TL;DR: A paper that introduces RIoT, a dataset aimed to address the scarcity of real-world IoT data.

Abstract: Despite the rapid proliferation of IoT devices, research progress is hindered by the lack of large-scale, publicly available datasets. Existing resources are often synthetic, short-lived, or overly simplistic, limiting their value for building realistic IoT systems. To address this gap, we present RIoT, a comprehensive real-world IoT dataset collected from six diverse building deployments over a cumulative six-year period. RIoT spans more than 200 sensors and over 10 million measurements across multiple modalities, including HVAC usage, energy consumption, and air quality, and captured at high temporal resolution with semantic and spatial annotations. By coupling scale with context, RIoT provides a realistic foundation for advancing IoT research, enabling the community to move from prototype-level studies toward robust, deployable systems.

Primary Area: datasets and benchmarks

Submission Number: 6343