Janmesh PatelRobotics software
PortfolioRobotics software engineerNew Jersey, USAOpen to roles

I write robotics software and test it on real robots.

I’m Janmesh Patel, a robotics software engineer in New Jersey. I work on perception, planning, control and the hardware they run on, mostly in C++17, Python, ROS 2 and Linux.

Highlights

  • Robot arm teleoperation at 100 Hz, 65 ms median camera-to-arm latency
  • Terrain-aware Nav2 costmap that cut median recoveries from 20 to 3
  • Integrated LiDAR, cameras and LTE on Clearpath mobile robots
  • Embedded work on STM32, FreeRTOS and Jetson, from board bring-up to ROS 2
Fig. 1: Top-down LiDAR view Simulated
1

Typical characteristics

Numbers I measured on my projects and at work.

ParameterConditionsValueUnitRef.
Control-loop rateFranka teleop on Jetson Orin Nano100HzNo. 01
Camera → arm latencyp50 / p9565 / 117msNo. 01
Control-loop periodp99, logged sessions10.1msNo. 01
Loop overrunsacross logged sessions0No. 01
Command → motionp50, after network fix and drive retune158 → 30msNo. 01
Median recoveriesterrain-aware vs default Nav220 → 3No. 02
Route timeterrain-aware vs default Nav2318 → 113sNo. 02
Assembly tolerancePCB and electromechanical assembly, Clearpath±0.0762mm
Competition winsRobocon, national finals50+
2

Projects

The problem, what I built, and the results for each one.

No. 01Robot controlOct 2026

Low-latency robot teleoperation & constrained IK

I built a 100 Hz teleoperation system that drives a simulated Franka arm from a tracked marker, running on a Jetson Orin Nano and Isaac Sim. The goal was low latency, and safe behavior when the network or robot state goes bad.

100HzControl loop
65msCamera → arm, p50
117msCamera → arm, p95
0Logged overruns
Before and afterBeforeAfter
MetricBeforeAfterUnit
Command → motion, p5015830ms
Average ping43.40.35ms
  • C++17
  • ROS 2
  • Jetson Orin Nano
  • RealSense D435
  • Isaac Sim
Read more
TELEOP LOOP · 100 HzILLUSTRATIVECameraRealSense D435Posemarker trackIKDLS + null-spaceSafetysupervisorArmFranka / Isaac65 ms p50 · 117 ms p95joint state · pose · safety state → synchronized episode loge-stop · stale state · limits
No. 02Autonomous navigationSep 2026

Terrain-aware navigation & vehicle safety stack

I wrote a Nav2 costmap layer that scores terrain from 3D LiDAR by slope, roughness and step height. Alongside it I built a vehicle safety arbiter and a simulated CAN drive interface that supports fault injection.

20→3Median recoveries
318→113sRoute time
−64%Time to goal
Before and afterBeforeAfter
MetricBeforeAfterUnit
Median recoveries203
Route time318113s
  • ROS 2
  • C++
  • Python
  • Nav2
  • CAN / DBC
Read more
TERRAIN COSTMAP · SLOPE + ROUGHNESS + STEPILLUSTRATIVESTARTCOSTsolid: terrain-aware pathdotted: flat-world path
3

Experience

Newest first.

  1. DMar 2026 to now

    Aetherion Robotics

    Founder · Robotics R&D

    My own self-funded R&D company. I work on autonomous ground vehicle concepts in Isaac Sim, covering navigation risk, motion planning and system testing. I also design the PCBs, bring up the boards and integrate the Jetson Orin Nano.

  2. CAug 2025 to Feb 2026

    Embedded & Robotics Systems

    Professional development

    Hands-on STM32/FreeRTOS and Embedded Linux projects: multitasking, SPI + DMA logging, telemetry, profiling, pthreads, PWM, signals and state machines.

  3. BJan 2024 to May 2025

    Clearpath Robotics by Rockwell Automation

    Robotics Specialist & R&D Technician

    Tested and debugged autonomous mobile robots running ROS 2 on Linux, tracking faults through software, firmware, comms, sensors, power and hardware. On the Husky A300 I integrated LiDAR, cameras, long-range Wi-Fi/LTE and wireless charging, and calibrated the LiDAR. I assembled parts to ±0.0762 mm and made assembly 50% faster with new tooling.

  4. ASep 2018 to Jun 2021

    The Roar Robotics

    Electronics & Programming Head · Robocon integration lead

    Led firmware and electronics for a team of 10+, using STM32, Arduino, Raspberry Pi and custom PCBs. We reached the Robocon national finals and won 50+ competitions.

4

Skills

Tools I use regularly.

Languages & systems

  • C++17
  • C
  • Python
  • Linux
  • Eigen
  • Multithreading
  • gtest / pytest
  • Latency & jitter profiling

Robotics & control

  • ROS 2
  • Nav2
  • Isaac Sim
  • Gazebo
  • Teleoperation
  • FK / Jacobian / DLS IK
  • Motion planning
  • SLAM
  • AMRs

Embedded & integration

  • Jetson Orin Nano
  • STM32
  • FreeRTOS
  • Embedded Linux
  • CAN / DBC
  • TCP / UDP
  • SPI · I²C · UART · DMA
  • PCB design & bring-up
  • Scope & logic analyzer

Sensors & data

  • RealSense D435
  • LiDAR
  • Cameras
  • OpenCV
  • Fast DDS
  • chrony
  • rosbag2 / MCAP

Work details

Location
New Jersey, USA
Work authorization
U.S. permanent resident, no sponsorship needed
Relocation
Open
Customer-site travel
Open

Education

2023
MEng, Systems and ComputingUniversity of Guelph
2021
BE, Mechatronics, Robotics & AutomationG H Patel College of Engineering & Technology