Five years of bench-level motion control on the SSi Mantra surgical arm, where I wrote the EtherCAT master. Alongside it, the tech stack of a Unitree G1 humanoid on a four-person programme: coordinator, safety path, navigation bridge, voice. 2,000+ commits across five platform generations.
A Unitree G1 with Inspire dexterous hands, 41 degrees of freedom, on a four-person programme where I am accountable for the tech stack, hardware accessories and software: 2,000+ commits across five platform generations. Monolithic proof-of-concept to a split-compute dual-Jetson platform, with a behavior-tree coordinator, a safety thread at the top of the scheduler, and a voice pipeline that never leaves the robot. Coordinator, safety, voice and the nav bridge all run on the real robot.
Priority arbitration Emergency > Teleop > Voice > Autonomous > Idle in a
BehaviorTree.CPP v4 coordinator. Arm joint commands go out on a 500Hz timer, the rate the DDS bus expects;
hand control publishes at 100Hz; processes exchange state over a hand-rolled lock-free seqlock shared-memory
interface, memory_order_acquire and all. Safety is its own 100Hz thread on a dedicated core, requesting SCHED_FIFO, watching thermal and battery thresholds. It never auto-damps, by design: the walking policy controls balance, so a damp from a side thread mid-walk causes the fall it exists to prevent. If the safety thread
dies, the coordinator releases the arms, opens the hands, raises the alert and leaves the damp to the operator.
That path used to report the robot as safe while its damp call did nothing. I found it and made it fail visibly.
Whisper large-v3-turbo → a custom fine-tuned 14B LLM → Kokoro TTS. The pipeline was originally squeezed onto a 16GB Orin NX (three models taking turns on one GPU behind a shared lock), and now runs on an AGX Orin 64GB with room to breathe. STT, TTS and VAD candidates each went through my own head-to-head shootout harness. A custom speculative-decoding path (draft model, batched verify, KV realignment) was written from scratch, benchmarked, and retired behind a flag when it lost on the number. In a design A/B, tag-routing beat LLM tool-calling by a measured 690ms per turn. Separately, I built a C++ speaker-ID pipeline that knows who is talking: Silero VAD → TitaNet embeddings, with temporal evidence gating.
The chain came up end to end on hardware: a goal set 1.5m out and 90° off heading, the robot aligned, walked 1.10m, and stopped within 10cm of it, eyes on, gantry-suspended. On the nav chain, my part is the bridge from planner to motors and everything guarding it, which is where the month actually went. Velocity and yaw-rate shaping under caps. Fail-closed on stale localization, measured at 1.53s from kill to emergency stop. No state feed, no drive: it publishes only zeros. Armed only by an explicit flag, disarmed by default, walked up a rung ladder where every step needed a human to say go. The planner itself (A*, elastic band, pure pursuit) is an existing stack I drive, not one I wrote.
A 500Hz DDS bridge (2ms timer) with cubic-Hermite interpolation, fed by MoveIt planners across OMPL, Pilz and CHOMP. Velocity scaling had no effect on execution in that pipeline, so I applied it in the interpolator by rescaling trajectory timestamps directly: a measured 10× execution-duration range from the same planned path.
Before/after numbers committed to git on an earlier generation of the voice stack. The wins came from KV-cache reuse, sentence-level LLM/TTS interleaving, and a GPU lock that hands each model the whole GPU on its turn, not bigger hardware. On today's stack, the same measure runs p50 2.82s, p95 4.82s over 40 live turns.
Five years on SSi Mantra, India's first indigenous surgical robot, a 12-DOF system. Motion control on its arm, validated at bench level: cascaded PID, gain scheduling and friction compensation on frameless-torque-motor joints, for smooth low-speed surgical motion. Hands deep in harmonic drives, ELMO and Ingenia servo controllers, Bode/Nichols on the bench for 7+ joints.
A pySOEM EtherCAT master for the arm: PDO mapping, the CiA 402 (DS402) drive state machine, bring-up and tuning across 7+ joints, and a bench commissioning flow that cut drive bring-up time by more than half. I wrote the master. Every byte on the bus accounted for.
Harnessing layouts, calibration procedures, repeatability and system-level safety compliance, the unglamorous work that lets a robot pass medical scrutiny. That habit carried over to the humanoid, where the safety path now fails visibly instead of reporting a success it never delivered.
At NUS in Dr. Marcelo Ang's lab, I designed and fabricated soft pneumatic actuators whose asymmetric bending is baked into the material itself: two elastomers cured as a viscosity gradient. Springer IAS 2022. An IEEE hexapod paper before that.
If any of this resonated (the humanoid, the bus, the bends without bones), write to me.