Most evenings, as we head home, the matatu drivers by the local footbridge wave at us. Just a quick nod or a smile, the kind of small gesture that slips quietly into your routine until one day, you realize it means something. That maybe you’re no longer just passing through.

This week, that feeling crept into more than just the journey home. Between testing low-cost electrodes, 3D printing wearable casts, and chasing down clean EMG signals, that sense of rhythm kept showing up – in soldered circuits, in shared coffees, and in all the in-between moments that are starting to feel a little more familiar.

Building the Active Cast One Snap at a Time

Our host-site project has taken shape – quite literally – over the last two weeks as we have worked with our colleague Waka at the design studio. The project focuses on designing a low-cost, wearable cast that can actively prevent muscle atrophy by stimulating muscles. The idea grew from conversations about the limitations of traditional immobilizing casts, especially after sports injuries, where muscle loss during healing can lengthen recovery time or limit function.

We began by debating the method of stimulation: PEMF (pulsed electromagnetic field) vs. electrical stimulation via electrodes. PEMF sounded great in theory, but sourcing components locally, cost-effectiveness, and simpler hardware pointed us toward electrical stimulation as the better fit for our context. So, we’ve been designing this circuit for the past two weeks. After finalizing the waveform we wanted (a biphasic square wave to avoid tissue damage), we coded the signal and started wiring the motor drivers and timing components. 

At one point, we accidentally burned through an Arduino board (RIP) – turns out, feeding it more than 22 volts is not a good idea! It also seemed like our multimeter wasn’t picking up the needed output voltage. Some troubleshooting later, we decided to go simple and test the output with LED lights instead… and it worked the whole time, the biphasic output was just too fast to be picked up by our multimeter. Classic. We’ve since cleaned up the design, confirmed the signal on an oscilloscope, and are looking into integrating this model into our cast design! 

Alongside electrical stimulation, we have also been building an EMG sensor (trying to capture real-time voltage signals from muscle contractions) so users can track muscle loss or gain over time. Early testing involved low-cost DIY electrodes made from washers, salt packets from the cafeteria, and hand soap instead of conductive gel (you can’t say we haven’t been getting creative!). At first, we struggled with inconsistent readings, and the op-amps available here weren’t ones we’d worked with before. 

Eventually, we rebuilt our differential amplifier, fine-tuned our reference voltage, and celebrated when we finally got readable signals on the oscilloscope!!

We then moved to designing the physical cast. Our first prototype was a flexible, single-piece sleeve with built-in slots for electrode placement meant to accommodate a variety of hand sizes. But once printed, the structure was too thin to offer any real support, and the electrode holes warped easily. So we pivoted.

We redesigned the cast into a two-part model that snaps together, giving it more rigidity and making it easier to print with more precise dimensions. This version has honeycomb-shaped inserts for electrodes and added thickness along the wrist to stabilize weak areas. It’s still moldable, but should hold up as we do further testing. We’re now going to start thinking about integrating the EMG sensor, the electrical stimulation circuit, and the cast together as one model. 

From Sound to Signal: Troubleshooting the Digital Fetoscope

In parallel with our active cast work, we’ve also been steadily building and refining the digital fetoscope. It’s been a process – writing code, testing filters, debugging hardware – but we’re starting to see clear, consistent signals.

We started troubleshooting last week with the basics: looking into the microphone, op-amp amplification, and a bandpass filter tuned to around 20–200 Hz. The first few days of coding were a lot of long hours. Our base Arduino code was noisy, and attempts to filter it either flattened the peaks or lagged too far behind to be useful. The turning point came when we started experimenting with Hilbert transforms in MATLAB, which helped us visualize the signal’s energy across both time and frequency. The heartbeats were starting to stand out more clearly.

Once we had that working offline, we started translating parts of the logic back into Arduino-compatible code. On Friday, we finally hit a breakthrough: the fetoscope began detecting beats per minute in real-time. The bandpass filter was comparatively stable, the peak detection consistent, and we could finally see those satisfying spikes when we played heartbeat audio near the mic. Progress!

Next up, we’re trying to re-incorporate a wavelet transform to denoise the signal more precisely, now that the base is stable. The hope is to preserve the heartbeat peaks while scrubbing out ambient and mechanical noise, especially from the mic, which keeps mysteriously losing sensitivity. We’re now on the hunt for a mic that can handle a wider frequency range – easier said than sourced, especially low cost. But hey, I love a challenge. Add it to the list of things I’ll be thinking about at 2 a.m. 

Back in the Field: Device Testing + Meetings

Last Wednesday, we delved deeper into our work with the Rice360 devices we brought with us, including the Endometrial Biopsy (EMB) trainer I helped develop back in Houston! It was surreal unboxing it here in Nairobi, where we’ll be testing it with clinicians for real-time feedback. We assembled the magnetic components, checked each tissue insert, and finalized the evaluation surveys for both this device and the neonatal gastroschisis bag.

Later that week, we met with Dr. George Okello, Rice360’s country lead, and his wonderful team. We coordinated clinician visits and feedback sessions with OBGYNs and pediatric surgeons. Sitting in that meeting, I had a moment. This is what I’ve always dreamed of: traveling, learning, and building solutions with real-world impact. And now, I’m in a room where it’s actually happening.

After the meeting, we took a short walk to the Java House down our street – a spot that’s quickly become part of our routine. Over spiced Malindi macchiatos (giving my homemade chai a run for its money!) and a Nairobi breeze strong enough to scatter our napkins, we caught our breath. Everyone was chatting, laughing, and checking in. Our days no longer feel new or uncertain. The same familiarity as the drivers on the footbridge seems to creep in here as well. 

Of Lions and the Coast

Over the weekend, we visited Nairobi National Park, where I saw lions nap beneath speeding trains and ostriches share space with the city skyline. It was a wild juxtaposition, and somehow, it made perfect sense. The edge between city and savanna felt thin, like a reminder that nature and structure can (and do) exist side by side. I’ve found that this feeling echoes throughout Nairobi itself. The city’s architecture feels like freedom: colorful, varied, and full of personality. It’s a city that doesn’t try to look the same, and it’s all the more alive because of it. This energy pulses through places like the Maasai Market – a rotating open-air bazaar filled with vibrant textiles, hand-carved jewelry, and bold paintings. Just walking through it feels like stepping into a palette of Nairobi’s creative spirit.

And just when I thought I’d seen the full range of contrast and color, we arrived in Mombasa.

The plan was simple: take a train to the coast for a weekend of rest and exploration. The reality? All the train tickets were sold out. That wasn’t going to stop us, though. After a sleepy 5:30 a.m. start, we pivoted fast, rerouting through flight bookings at the station and somehow still finding time to stop by the Giraffe Centre, where we fed a three-month-old baby (yes, it was adorable).

With a couple of hours still remaining before our flight, we found ourselves sitting under sun-dappled trees at Cultiva, a farm-to-table spot with amazing brunch and a few wandering cats for company. My aunt, who worked in Nairobi for a year, has been giving us the best food recs!! And then – one flight delay and a 90-minute drive from the Mombasa airport later – we finally made it to Diani Beach just in time for the Summer Tides Music Festival.

That evening, we met up with our fellow Rice CCL interns and local friends, caught some incredible music right on the sand, and found new artists to obsess over (Mutoriah and Nikita Kering – highly recommend).

The next morning, we took the ferry into Old Town Mombasa, which was a maze of coral stone buildings, carved wooden doors, and ocean air thick with salt and spice. As we wandered through the narrow streets, I struck up a conversation with a computer science student working at Nyali Beach for the summer. A few steps later, just outside Fort Jesus (a 16th century Portuguese Fort), a welding technician was crouched beside the road, carefully shaping metal into ornate hinges for a carved wooden gate.

There were sparks flying in the open air, the hum of tools blending with the call to prayer in the distance. It made me smile – we’d had a go at welding joints during the summer program, but seeing that level of precision in action was something else entirely.

Closing the Loop

These past two weeks, our fetoscope picked up real beats, our circuits pulsed in square waves, and we squeezed in some adventures along the way. We made progress in code, in conversation, and in the quiet moments in between.

Each evening, we cross that same footbridge – the one where the matatu drivers now nod as we pass –  and return to a place where our neighbors smile and wave. A place that feels a little more like home. What began as unfamiliar is starting to take shape in the form of the patterns we’re slowly becoming part of. One familiar moment at a time. 

Signing off from our little corner of Nairobi,
Saumya 🏙️⚙️