Articles
A Conversation with Nikita Joshi: Using IoT and AI to build a more proactive and accessible healthcare future
Full citation in the ACM Digital Library 
| PDF 
Ubiquity
Volume 2025, Number October (2025), Pages 1-5
Innovation Leaders: A Conversation with Nikita Joshi: Using IoT and AI to build a more proactive and accessible healthcare future
Bushra Anjum
DOI: 10.1145/3772189
In this interview, Ubiquity senior editor Dr. Bushra Anjum speaks with Dr. Nikita Joshi about the evolving role of computing in healthcare. The discussion traces Dr. Joshi's journey from IoT-based remote patient monitoring to AI-driven mental health prediction using data from smartphones and wearable sensors. It highlights how her work integrates IoT, fog computing, and machine learning to enable proactive, real-time health monitoring and explores the ethical and practical dimensions of using AI to make healthcare more accessible and preventive.
Dr. Nikita Joshi is a researcher and educator specializing in IoT-enabled healthcare systems and machine learning for mental health prediction. Her Ph.D. research focused on remote patient monitoring, where she explored how IoT sensors and fog-cloud computing can enhance real-time healthcare analysis. Over the years, she has worked on optimizing resource allocation in IoT-based healthcare to ensure efficient and low-latency processing of patient data.
Post-pandemic, she transitioned to a more specific application—mental health prediction—by leveraging smartphone and wearable sensor data. Her current work focuses on applying AI/ML techniques to detect early signs of mental health issues, providing individuals with actionable insights to take preventive measures. Driven by a passion for real-world problem-solving, she envisions technology bridging the gap between healthcare accessibility and proactive intervention. She can be contacted via her LinkedIn page.
What is your main concern about the future of computing in proactive healthcare monitoring?
One of the most pressing concerns in healthcare computing is the lack of continuous and proactive patient monitoring, especially for chronic disease management and elderly care. Traditional healthcare relies on periodic checkups, making it difficult to detect health issues before they escalate. Many patients require constant observation, yet medical staff shortages and hospital overcrowding make real-time monitoring impractical. This creates a growing demand for smart, automated healthcare solutions—a need recognized in the WHO's "Global Strategy on Digital Health 2020–2025," which highlights the importance of scalable technologies to support an aging population, address workforce shortages, and manage the increasing burden of chronic diseases.
The internet of things (IoT) has revolutionized healthcare by enabling real-time patient monitoring through wearable devices and smart sensors. These devices collect vital health metrics such as heart rate, activity levels, sleep patterns, and oxygen saturation, allowing patients to be monitored remotely. This approach is particularly beneficial for elderly individuals, post-operative patients, and those managing chronic conditions, reducing hospital visits while ensuring timely intervention.
However, IoT devices generate massive amounts of data, which cannot be processed locally due to their limited computing power and energy constraints. To address this, cloud computing has been widely adopted, but it introduces latency and network dependency issues, making it unsuitable for critical real-time applications. Fog computing—a decentralized computing model that processes data at or near the source instead of sending it all the way to the cloud—has gained importance. In the context of healthcare, this means that critical patient data collected from wearable sensors can be analyzed locally (e.g., in nearby servers or edge devices), enabling faster response times and reducing dependence on distant cloud servers. Fog computing is especially useful when latency is critical, such as when detecting a sudden drop in oxygen levels or a cardiac irregularity.
My research is dedicated to leveraging IoT, fog computing, and AI-driven analytics to create intelligent, real-time healthcare monitoring systems. By integrating machine learning algorithms, these systems can predict potential health risks and enable early intervention, shifting healthcare from a reactive to a proactive model. While my initial focus was on physical health monitoring, I later expanded my research into mental health assessment, where AI can analyze behavioral data from wearables to detect signs of psychological distress.
By developing scalable, AI-powered healthcare solutions, my goal is to make continuous, accessible, and intelligent health monitoring a reality, ensuring that no patient goes unnoticed.
What prompted you to expand your focus from IoT and fog computing to your current interest in mental health analytics?
My passion for technology-driven healthcare stems from my Ph.D. research on IoT-based remote patient monitoring. During this time, I observed that many patients needed continuous health tracking, but the traditional approach of in-person checkups was impractical and inefficient. This motivated me to explore how IoT devices could automate patient monitoring, sending real-time health data to cloud-based systems for analysis. However, I soon realized that cloud computing alone was not sufficient, as latency and data overload posed significant challenges. This led me to study fog computing as a way to process data closer to the patient, ensuring low-latency, real-time decision-making.
However, my perspective on healthcare technology broadened when I saw the mental health crisis intensify during the COVID-19 pandemic. It became clear that mental well-being is just as critical as physical health, yet current healthcare systems lack objective tools to assess and track mental health conditions. I realized that smartphones and wearable sensors could passively collect behavioral and physiological data, such as sleep patterns, heart rate variability, and social activity, which are strong indicators of mental health.
Given my background in AI and IoT, I saw an opportunity to explore how machine learning could support mental health prediction by analyzing real-time behavioral data. Traditional self-reported assessments, while important, often depend on subjective inputs and are not continuous. In contrast, recent research suggests that AI models can detect early signs of mental distress by identifying patterns in passively collected data, such as changes in sleep habits, reduced physical activity, or irregular smartphone usage. While these models are not a replacement for clinical diagnosis, they offer a promising complementary approach by providing continuous, non-intrusive insights that could help individuals and healthcare providers intervene earlier.
Throughout my journey, I have remained committed to using technology to solve real-world healthcare challenges. From remote patient monitoring to mental health prediction, my work has been shaped by the belief that computing technologies can make healthcare more proactive, accessible, and efficient.
How do you see collaboration between AI researchers and healthcare professionals shaping the future of mental health technology?
Currently, I am leading a research project focused on AI-powered mental health prediction using smartphone and wearable sensor data. In our initial phase, we have developed and tested a prototype model that collects behavioral data (e.g., sleep patterns, phone usage) and applies machine learning algorithms to detect patterns associated with stress, anxiety, or depression.
Our early experiments using models like K-nearest neighbors (KNN) have shown promising accuracy—with results reaching showing good accuracy in classifying mental health risk levels. We are now working on expanding the dataset and validating our models across a more diverse user base to improve generalizability and reliability.
While the current focus is on data collection and predictive modeling, our future plan includes integrating real-time feedback into a mobile application. This app will aim to provide users with personalized insights and alerts, potentially guiding them to take preventive actions or seek professional help.
Ultimately, the goal is to build a scalable and ethical AI-driven support system that complements clinical approaches and brings mental health tracking into everyday life through familiar technologies like smartphones and wearables.
In the future, I aim to expand this research into personalized intervention systems, where AI can recommend coping strategies or connect individuals to mental health professionals based on real-time behavioral analysis. The goal is to create a seamless, AI-driven support system that empowers individuals to manage their mental well-being proactively.
I am passionate about collaborating with researchers, healthcare professionals, and technologists to advance AI and IoT-driven healthcare solutions. If you are working on AI applications in healthcare, IoT-based health monitoring, or mental health analytics, I invite you to connect. Let's work together to build intelligent, real-time healthcare systems that enhance patient care and well-being.
Author
Dr. Bushra Anjum is a strategic data science and AI leader with more than a decade of experience building high-impact data and AI solutions. With a doctorate in computer science, she possesses deep expertise in stochastic modeling, machine learning, LLM development and AIOps. Throughout her career, she has successfully architected and launched innovative data-driven solutions in the technology, healthcare and education sectors.
2025 Copyright held by the Owner/Author.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2025 ACM, Inc.
COMMENTS