The posting
About Origin
Origin (previously 10xConstruction) is building general-purpose autonomous robots for US construction to tackle rising costs, safety risks, and labour shortages. Our modular, multi-trade platform combines purpose-built hardware with real-time site intelligence to navigate complex environments and execute tasks with precision. Trained in high-fidelity simulation and already deployed on live sites, our robots deliver 5x faster execution, 250%+ margin expansion, and significant cost savings. Join India's most talent-dense robotics team consisting of individuals from IITs, Stanford, UCLA, etc.
About the role
Robotic drywall finishing is a hard, unsolved problem — and you'll be in the thick of solving it. This role sits at the intersection of three things: building the physical setups we experiment on, running the experiments that tell us what actually works, and automating those experiments so we can run far more of them. You'll stand up test rigs and instrumentation, design and execute DOEs on the robotic finishing process, and then write the ROS 2 and Python tooling that turns a manual trial into a repeatable, scripted protocol. It is deliberately interdisciplinary — mechanical assembly, embedded systems and sensors, robot software, and experimental method all in the same week. We're looking for someone whose experience is broad rather than narrow: two years spent solving a variety of problems across several projects teaches you more of what this role needs than two years spent deep inside a single one. If you think like an experimentalist, are comfortable with a wrench in one hand and a terminal in the other, and want your work to visibly move the system forward, this role is for you.
Key Responsibilities
Build and maintain experimental setups
- Design, build, instrument, and maintain test rigs, fixtures, and experimental setups for robotic drywall finishing trials
- Integrate sensors, end-of-arm (EOA) tooling, actuators, and data acquisition hardware into working setups; own wiring, mounting, calibration, and day-to-day upkeep
- Debug setups end to end across mechanical, electrical, embedded, and software layers, and keep them in a known-good, repeatable state
- Improve setup design over time so experiments get faster, cleaner, and easier to reproduce
Design and run experiments
- Design and run controlled experiments on robotic drywall finishing processes using hypothesis-driven and DOE-based approaches
- Identify and isolate independent, dependent, controlled, and confounding variables across physical trials
- Maintain repeatability and reproducibility discipline in measurement workflows; know the difference between a real effect and setup noise
- Analyse cross-system interactions across the robot arm, end-effector, sensors, surface material, and motion parameters to support debugging and process improvement
- Document observations rigorously and convert trial results into clear experiment logs, test reports, and result summaries that engineering decisions can be based on
Automate test protocols
- Automate test and measurement protocols using ROS 2 and Python to increase the rate of experimentation
- Script robot motion sequences, parameter sweeps, and data capture so trials run consistently with minimal manual intervention
- Build tooling for automated data collection, logging, and first-pass analysis; reduce the manual effort between "run the trial" and "read the result"
- Work with the robotics team on deep data collection from EOA systems with advanced sensors — from wiring and drivers through to structured, analysable datasets
Required Qualifications and Skills
- B.E./B.Tech/M.E./M.Tech in Mechanical, Mechatronics, Robotics, Electrical/Electronics, Instrumentation, Industrial, Aerospace, or Systems Engineering (candidates from Applied Sciences / Experimental Physics with strong hands-on engineering exposure may also be considered)
- ~2 years of hands-on experience in the robotics and automation field, across a variety of solutioning-driven projects. Breadth matters here: we value someone who has shipped several different builds, integrations, or test systems over someone who has spent the same period on one narrow scope
- Demonstrated experimentation experience in a lab, R&D, or test-engineering environment — designing trials, not just executing someone else's checklist
- Working knowledge of ROS 2 (nodes, topics, launch files, basic packages) and comfort writing production-quality-enough Python for automation, data capture, and analysis
- Basics of embedded systems — microcontrollers, sensor interfacing, common protocols (UART/I2C/SPI/CAN), and the ability to read a datasheet and get a sensor talking
- Hands-on competence with electromechanical assemblies, sensors, instrumentation, wiring, and calibration
- Understanding of experimental methods, statistics/DOE, control systems, and system design through coursework, research, or industry work
- Ability to work with engineering data from physical trials and perform statistical analysis and interpretation
- Strong documentation skills — structured test reports, experiment logs, and result summaries
- Genuine appetite to learn across disciplines, and comfort working on an unsolved problem where the right answer isn't known yet
Preferred Experience
- Experience with industrial or collaborative robot arms — UR, KUKA, ABB, Fanuc, or similar — including motion programming and integration
- Prior work with imaging-based experimentation setups (computer vision, machine vision, optical or laser-based measurement)
- Experience designing test setups, instrumentation chains, or validation workflows from scratch
- Force/torque sensing, IMUs, depth cameras, encoders, or other advanced sensing integrated into a working data pipeline
- Familiarity with data acquisition systems, real-time logging, or time-synchronised multi-sensor capture
- Comfort with Excel/Google Sheets, Python (NumPy/pandas/matplotlib), or MATLAB for experiment logging, analysis, and plotting
- Exposure to CAD tools (SolidWorks / Fusion 360 / Onshape) for designing fixtures and brackets
- Version control (Git) and documentation tools (Notion, Confluence, Jira, or Google Docs)
- Exposure to simulation environments (Gazebo, Isaac Sim) or process/materials domains such as coatings, surface finishing, or abrasives
Why this role
You will own the loop between a physical question and a defensible answer — building the rig, running the experiment, and automating it so the next hundred runs are cheap. The scope is broad by design, the feedback is fast, and the results go straight into a robot working on live construction sites.
