Zephyr RTOS firmware
About 15,000 lines of Zephyr C on nRF52840 (Arm Cortex-M4F): sensor sampling and scheduling, store-and-forward buffering, camera capture, and drivers for I2C, SPI, UART, RS-485, and SDI-12.
Nurzaman Ahmed
# Engineering research scientist and embedded systems lead spanning Wi-Fi MAC, Zephyr RTOS firmware, cellular IoT, interoperable AI, edge-cloud, and digital agriculture.
60+:: Scholarly works & chapters
13+ yrs:: Connected systems research
4:: Patents granted or filed
STL / USA:: Current location
Profile
I design wireless protocols and intelligent systems that turn real-world data into reliable infrastructure for discovery and action.
My work moves from IEEE 802.11 MAC design and Linux wireless driver implementation to Wi-Fi HaLow field networks, edge intelligence, plant phenotyping, and interoperable agent systems. The constant is practical: research should survive real constraints and become something people can use.
At Agrela Ecosystems, I lead systems engineering for PheNode, a solar-powered field sensing platform: Zephyr RTOS firmware on nRF52840, LTE-M/NB-IoT cellular and LoRa connectivity, secure over-the-air updates, and the fleet cloud behind it.
Selected systems
view all repositories →| ID | Project | Domain | Status | |
|---|---|---|---|---|
| sys-00 | PheNode · Agrela EcosystemsSolar-powered field sensing platform: Zephyr RTOS firmware on nRF52840, LTE-M/NB-IoT cellular and LoRa sensor links, end-to-end over-the-air updates, and a fleet cloud with device telemetry. | embedded / cellular IoT | [commercial] | ↗ |
| sys-01 | OpenWrt ath9k 2C MACA driver-level TDMA-style MAC implementation in Linux ath9k/mac80211, with modified C driver paths, raw-packet tools, and OpenWrt/Linux testbed evaluation. | Wi-Fi MAC / driver | [public] | ↗ |
| sys-02 | Agent Skills Interoperability LabPortable patterns for distributed task routing, MCP/OpenAPI tool use, tracing, evaluation, and prompt-injection resistance. | agent systems | [public] | ↗ |
| sys-03 | IoT Learning GameA game-oriented IoT learning environment with real-time sensor data visualization and hands-on activities. | IoT education | [public] | ↗ |
| sys-04 | Internet of Agents for AgricultureA modular agriculture automation framework connecting AI agents, IoT devices, field observations, and action workflows. | agentic IoT | [public] | ↗ |
| sys-05 | SubterraAIA human-in-the-loop platform for image-based root phenotyping, versioned model retraining, and GWAS-ready analysis. | root phenotyping | [public] | ↗ |
| sys-06 | FieldDockAn integrated field platform linking sensing, edge computing, drone operations, APIs, and crop-breeding workflows. | digital agriculture | [deployed] | ↗ |
Embedded & cellular IoT
phenode.com →# Lead Systems Engineer at Agrela Ecosystems, building PheNode: solar-powered field sensing from the sensor bus to the cloud.
Cellular & LoRa connectivity
LTE-M/NB-IoT through Blues Notecard with per-device link telemetry (RSSI, SINR, access technology), LoRa sensor links on SX1262, and BLE, with data buffered through network outages.
Secure over-the-air updates
An end-to-end OTA pipeline: cloud to gateway over cellular, then on to wireless sensors over AES-CMAC-authenticated LoRa, with SHA-256 verification and MCUboot rollback.
Bring-up, power & fleet
Board bring-up checklists, Otii power profiling, oscilloscope bus debugging, and a fleet cloud with remote configuration, watchdog recovery, and device health dashboards.
Research focus
# One systems question at every scale: how do we make intelligence reliable, interoperable, and useful where data is born?
Wi-Fi MAC & embedded wireless
IEEE 802.11 MAC and Wi-Fi HaLow architecture; QoS, scheduling, relay, and multi-hop design; Linux/OpenWrt mac80211 and ath9k driver work; packet-level debugging, BLE, and Zephyr RTOS.
Interoperable AI systems
Agent workflows, tool contracts, observability, evaluation, and the infrastructure that makes model behavior inspectable.
Edge-to-cloud intelligence
Reliable architectures that move computation across constrained devices, programmable networks, and cloud analytics.
Digital agriculture
Field-ready sensing, UAV imaging, plant phenotyping, and decision systems for productive, climate-smart agriculture.
Selected publications
full archive →Research trajectory
Agrela Ecosystems · Saint Louis · PheNode
Donald Danforth Plant Science Center · Saint Louis
Dartmouth College · Computer Science
Indian Institute of Science · ECE
Indian Institute of Technology Kharagpur · CSE
North-Eastern Hill University
Open channel
# Available for research conversations and engineering collaborations around Wi-Fi MAC, embedded firmware and RTOS, cellular IoT, wireless drivers, agentic and interoperable AI, edge computing, and digital agriculture.