Intelligent Sensing to Inform and Learn (InSTIL): A Scalable and Governance-Aware Platform for Universal, Smartphone-Based Digital Phenotyping for Research and Clinical Applications

Scott Barnett, Kit Huckvale, Helen Christensen, Svetha Venkatesh, Kon Mouzakis, Rajesh Vasa

Abstract

In this viewpoint we describe the architecture of, and design rationale for, a new software platform designed to support the conduct of digital phenotyping research studies. These studies seek to collect passive and active sensor signals from participants' smartphones for the purposes of modelling and predicting health outcomes, with a specific focus on mental health. We also highlight features of the current research landscape that recommend the coordinated development of such platforms, including the significant technical and resource costs of development, and we identify specific considerations relevant to the design of platforms for digital phenotyping. In addition, we describe trade-offs relating to data quality and completeness versus the experience for patients and public users who consent to their devices being used to collect data. We summarize distinctive features of the resulting platform, InSTIL (Intelligent Sensing to Inform and Learn), which includes universal (ie, cross-platform) support for both iOS and Android devices and privacy-preserving mechanisms which, by default, collect only anonymized participant data. We conclude with a discussion of recommendations for future work arising from learning during the development of the platform. The development of the InSTIL platform is a key step towards our research vision of a population-scale, international, digital phenotyping bank. With suitable adoption, the platform will aggregate signals from large numbers of participants and large numbers of research studies to support modelling and machine learning analyses focused on the prediction of mental illness onset and disease trajectories.

Introduction

Digital phenotyping—the use of smartphone sensors to track and predict health outcomes—holds immense promise for mental health. By passively collecting data such as GPS mobility, sleep patterns via accelerometry, or social activity, researchers can identify the earliest signs of relapse or illness. However, the field currently faces significant hurdles, including technical fragmentation, the "iOS Gap" caused by strict backgrounding restrictions, and complex governance requirements for sensitive behavioral data.

InSTIL (Intelligent Sensing to Inform and Learn) is a new, scalable, cross-platform infrastructure designed to standardize this process. By providing a unified SDK for both iOS and Android and a privacy-first cloud backend, it lowers the technical barrier for researchers and paves the way for a population-scale "international digital phenotype bank."

The InSTIL Platform Architecture

The InSTIL platform is designed to be "governance-aware" and scalable, managing the entire lifecycle of a research study from design to data extraction.

The InSTIL platform workflow, illustrating the lifecycle of a digital phenotyping study from design to data extraction.

Key technical components include:

  • Cross-Platform SDK: Native libraries for iOS (Swift) and Android (Java/Kotlin). To solve the iOS backgrounding issue, InSTIL uses a communication protocol that relies on silent background notifications to periodically "wake" the app for data collection.
  • Cloud-Native Backend: Built on Kubernetes for scalability, the backend utilizes Apache Kafka for high-throughput data pipelines, Elasticsearch for indexing, and PostgreSQL for persistent storage.
  • Agnostic Ingestion Pipeline: The server is designed to be "type-agnostic," allowing researchers to add new sensors (like wearables) without rewriting backend code.
  • Privacy Protocol: The platform is "anonymous by default." It uses a custom enrollment protocol where the platform only sees a unique ID. Researchers maintain their own separate "linkage keys" for clinical follow-up, ensuring cloud data remains de-identified.

Design Considerations and Governance

The development of InSTIL was guided by several core principles aimed at ensuring equity and data integrity:

  • Universal Access as Equity: Supporting iOS is viewed as an ethical necessity to ensure "equity of opportunity" for research participants across different socioeconomic backgrounds, preventing the data biases common in Android-only studies.
  • The UX/Data Trade-off: High-resolution sampling provides better data but impacts battery life. InSTIL allows researchers to dynamically tune these parameters via a web interface without requiring an app update, helping to mitigate "app churn."
  • Governance as a Feature: The platform acts as a formal store for IRB approvals. Data collection cannot commence until the appropriate ethics documentation is uploaded to the Experiment Manager.

Implementation and Future Directions

While primarily a system architecture paper, InSTIL has already been deployed in the Future Proofing Study, a large-scale school-based depression prevention trial. The architecture is designed to support tens of thousands of concurrent users with robust synchronization logic to handle intermittent connectivity.

Future work will focus on real-time data validation (e.g., checking audio sample quality on the fly) and addressing the intrinsic identifiability of GPS data through methods like "temporal escrow" and coordinate remapping. While a shared platform reduces development costs, the long-term sustainability of operational costs for a global phenotype bank remains a key area for future modeling.

Summary updated

Abstract

In this viewpoint we describe the architecture of, and design rationale for, a new software platform designed to support the conduct of digital phenotyping research studies. These studies seek to collect passive and active sensor signals from participants' smartphones for the purposes of modelling and predicting health outcomes, with a specific focus on mental health. We also highlight features of the current research landscape that recommend the coordinated development of such platforms, including the significant technical and resource costs of development, and we identify specific considerations relevant to the design of platforms for digital phenotyping. In addition, we describe trade-offs relating to data quality and completeness versus the experience for patients and public users who consent to their devices being used to collect data. We summarize distinctive features of the resulting platform, InSTIL (Intelligent Sensing to Inform and Learn), which includes universal (ie, cross-platform) support for both iOS and Android devices and privacy-preserving mechanisms which, by default, collect only anonymized participant data. We conclude with a discussion of recommendations for future work arising from learning during the development of the platform. The development of the InSTIL platform is a key step towards our research vision of a population-scale, international, digital phenotyping bank. With suitable adoption, the platform will aggregate signals from large numbers of participants and large numbers of research studies to support modelling and machine learning analyses focused on the prediction of mental illness onset and disease trajectories.
BibTeX
@article{barnett2019intelligent,
  author = {Scott Barnett and Kit Huckvale and Helen Christensen and Svetha Venkatesh and Kon Mouzakis and Rajesh Vasa},
  title = {{Intelligent Sensing to Inform and Learn (InSTIL): A Scalable and Governance-Aware Platform for Universal, Smartphone-Based Digital Phenotyping for Research and Clinical Applications}},
  journal = {Journal of Medical Internet Research},
  year = {2019},
  doi = {10.2196/16399},
}
Figure