Zohaib Akhtar MD MPH
In 2018, my team and I aimed to solve one of the most pervasive yet often overlooked challenges in mental health: isolation. At its heart, isolation isn’t just about being alone—it’s about lacking meaningful connections. In high-pressure environments like universities, where people are juggling academic stress, cultural transitions, and personal challenges, this sense of disconnection can lead to anxiety, depression, and other mental health struggles.
The idea for Dropby was to address this problem by building a system that didn’t just treat mental health issues but worked to prevent them by fostering meaningful, real-world connections. Over time, it grew into a massive project involving collaborations with the Mental Health Department, Biostatistics, Health Informatics, and Engineering teams. We worked with postdoctoral researchers from multiple disciplines and had the invaluable support of founder of Plone, the communication system used by the FBI and CIA, who helped with concept development. We built a comprehensive system designed to connect people, improve well-being, and gather never explored data points to understand the links between behavior and health.
A tool and public health
Dropby was more than a social platform. It was a privacy-first public health system built around two goals:
-
To foster meaningful real-world connections that reduce isolation and improve mental health.
-
To gather and analyze data that could reveal patterns between social interactions, physical activity, and mental health outcomes.
The idea was simple but powerful: By connecting people through shared activities and integrating health insights, we could tackle isolation while building a better understanding of well-being.
How it worked
Our system featured a mobile application designed to connect people in an engaging ways. It leveraged maps, physical movement, and proximity to other devices to create opportunities for users to find and join activities happening in their area. The app provided an intuitive interface where users could select and participate in events based on their interests, fostering real-time interactions. The system highlights includes:
1. Connecting Through Activities
The platform focused on bringing people together based on shared interests. Users could choose activities like cycling, coffee meetups, or study groups, and the system matched them with others nearby who shared the same preferences. Early versions supported one-on-one connections, while later updates added group activities and an interactive map that displayed real-time events. This feature made both planned and spontaneous interactions easy and enjoyable.
2. Privacy-First Design
Privacy and safety were central to the platform’s design. Profiles were verified through the university system to ensure users were real and accountable. The system allowed users to see only up to 10 nearby connections within a 1-mile radius, creating a balance between connection opportunities and privacy. By leveraging Bluetooth Low Energy (BLE) instead of GPS, the platform provided precise, localized connections while protecting users’ location data, ensuring trust without compromising functionality.
3. Real-Time Features
One of the most innovative aspects of the platform was its real-time functionality. Users could see ongoing activities on an interactive map, join events instantly, and communicate through a built-in messaging system. Dynamic scheduling features matched users based on their availability and preferences, making social interactions seamless and spontaneous.
4. Data-Driven Insights
Beyond creating connections, the system aimed to understand the impact of social activities on health. By integrating data from wearable devices—such as heart rate variability, sleep patterns, and activity levels—with in-app behavior, the platform provided personalized insights into how social interactions influenced mental and physical well-being. These insights also supported broader public health research, linking user behavior to measurable health outcomes.
A collaborative effort
Dropby was the result of a truly multidisciplinary collaboration, bringing together expertise from a diverse range of fields. Mental health professionals, biostatisticians, health informatics researchers, engineers, and business strategists worked side by side to tackle both technical and human challenges. This integration of perspectives ensured that Dropby was not only technologically advanced but also deeply rooted in evidence-based mental health strategies, with a clear path toward long-term sustainability.
The collaboration played a pivotal role in navigating the complexities of software and medical device development, addressing both regulatory and technical hurdles. Postdoctoral researchers from various departments brought cutting-edge insights to refine Dropby’s design and expand its capabilities. Faculty from the Bloomberg School of Public Health contributed to developing robust analytics to measure outcomes, while experts from the Carey Business School worked on creating a scalable and sustainable business model. This ensured that Dropby had the potential to grow beyond its pilot phase and become an impactful public health tool for diverse communities.
One of the most remarkable aspects of Dropby’s development was the active engagement of hundreds of students from Johns Hopkins University. These students didn’t just test the system—they were integral to its design. Their feedback helped shape key features, such as activity-centric group meetups and real-time event notifications, ensuring that the app was user-driven and addressed their real-world needs. By incorporating student voices, Dropby became a relevant and practical solution for the challenges its users faced.
This unique combination of expertise, creativity, and collaboration made Dropby much more than an app. It became a system designed by the very people it aimed to serve—a powerful, impactful tool for mental health and community building. Dropby stands as a testament to what can be achieved when diverse disciplines unite with a shared commitment to creating meaningful, user-focused innovation.
Technology challenges
Building Dropby’s networking system was both exciting and challenging. The app used a combination of Bluetooth Low Energy (BLE)and GPS-based mapping to connect users in real-time while keeping their privacy protected. GPS helped power a map feature that allowed users to see and join activities happening across a large area, like campus-wide events. Meanwhile, BLE was used for short-range, precise connections indoors, where GPS often struggles to work well. By combining these two technologies, Dropby aimed to create a system that was both accurate and scalable. However, making this work wasn’t easy.
One of the main challenges was power consumption. BLE is supposed to use less power than GPS, but in places with a lot of users—like crowded lecture halls or libraries—devices sometimes overheated, and their batteries drained quickly. This became a bigger issue when many devices tried to connect at the same time, which often caused the system to overload and stop working properly. This made it harder for Dropby to reliably connect users in these situations.
Another issue was accuracy. BLE didn’t always correctly estimate how close people were to each other. Sometimes it detected users who were far away (false positives), and other times it missed users who were nearby (false negatives). These errors affected the app’s ability to make connections that felt seamless and precise, especially in indoor spaces where accuracy was really important.
There were also challenges with getting Android and iPhone devices to work together. BLE behaves differently on each operating system, so creating a system that worked consistently across both required a lot of extra effort. This issue became even more obvious during the early development of COVID-19 contact tracing apps, which faced similar problems. It wasn’t until Apple and Google introduced new standards that these issues started to improve. Today, technologies like AirTags make BLE systems seem simple, but at the time Dropby was being built, overcoming these technical hurdles was a big achievement.
Despite these challenges, the combination of BLE and GPS in Dropby was a forward-thinking approach. BLE made it possible to create precise, short-range connections indoors, while GPS allowed users to see activities happening across larger areas. This mix helped Dropby maintain user privacy while offering a system that was scalable and practical. While it wasn’t perfect, the lessons learned from developing Dropby paved the way for improvements in public health technology and other tools that rely on proximity-based networking. It was a groundbreaking effort that pushed the limits of what was possible at the time.
Why Dropby was important.....and still is
Dropby began as a thoughtful response to a problem that can lurk beneath the surface for students in demanding environments: isolation. Early on, our research revealed how newcomers and international students often feel cut off, overwhelmed by cultural shifts, academic pressures, and challenges navigating university resources. It quickly became clear that isolation isn’t just about being physically apart—it’s a complex interplay of behavior, environment, and technology. To meet that complexity, we built a platform that went beyond a simple social app. Dropby quietly guided students toward real-life gatherings (study groups, coffee meetups, sunrise rides) using proximity sensing, intuitive maps, and strict privacy protections. By fusing what people did in the app with health insights from wearables (like heart rate, sleep, and activity), we offered personalized suggestions that linked meaningful connections directly to well-being, all while preserving trust through privacy‑first design.
However, when the COVID-19 pandemic hit, strict social distancing guidelines made the in-person gatherings at the heart of Dropby impossible, and the project had to be paused. Looking back, we recognize that AI could now power systems resilient to such disruptions, blending virtual and physical connections to maintain community even in isolation.
Today, artificial intelligence opens a new chapter for Dropby’s vision. Imagine a system that learns your rhythms and gently surfaces the meetups that brighten your day, whether that’s a late‑night brainstorming session or a brisk lunchtime jog. Behind the scenes, models trained on patterns of past behavior can forecast when someone might be drifting into isolation and automatically deliver a friendly prompt or an event invitation right when it matters. On‑device learning and privacy‑enhancing techniques (such as federated learning and differential privacy) ensure your personal journey remains yours while collective insights inform smarter matches for everyone.
AI can also sharpen the way Dropby senses the world. By blending signals from Bluetooth, GPS, and wearable sensors through advanced filtering algorithms, the platform can more accurately understand who is truly nearby and what context they’re in, avoiding false connections in crowded hallways or noisy lecture halls. At the same time, machine‑driven anomaly detection can flag any harassment or unusual patterns, helping keep interactions safe and welcoming.
Looking forward, these foundations can evolve further. Continual learning algorithms will let the system adapt as campus cultures shift, taking into account not just steps and sleep but nutrition, stress markers, and even emotional cues from in‑app conversations. Augmented reality overlays could guide you to pop‑up wellness events, and closed‑loop AI pipelines will refine programming based on actual health outcomes, bringing together social engagement, technology, and data into a seamless experience. In this way, Dropby becomes more than a connector—it grows into an empathetic, proactive companion focused on keeping communities vibrant and individuals thriving.
AI-driven analytics can take this a step further by examining both community trends and individual journeys to power personalized goals. By analyzing data (from how often someone joins a study group to shifts in sleep or heart rate) intelligent models can suggest tailored targets, like a weekly social-connect quota or balanced activity-rest ratios, aligned with each person’s needs. Interactive dashboards and predictive insights let users and support teams monitor progress against clear health outcome objectives, while automatic feedback loops continuously refine these goals. This way, Dropby doesn’t just spark connections—it guides every individual toward measurable improvements in mental and physical well-being. I’m excited to collaborate with anyone who shares this dream of using AI to build health‑focused social systems that truly move the needle on well‑being.
