Robotics · Mechatronics · Systems Engineering

Lyle
Kosinski

Ten years engineering the systems that move semiconductor chips across the world — now building robots that keep ships clean.

Lyle Kosinski
10+Years Experience
1Patent Held
MSRobotics & Mechatronics
99thPercentile — Spatial Perception

Background

About

I'm a robotics and systems engineer with a decade of experience designing, integrating, and qualifying robotic semiconductor wafer-handling systems at Applied Materials — coordinating across mechanical, firmware, and software disciplines to deliver reliable, production-grade automation to fabs worldwide.

I hold an MS in Robotics & Mechatronic Systems from Santa Clara University and a BS in Mechanical Engineering (mechatronics concentration) from San Jose State. As a designated Engineer of Authority across multiple business units, I owned the full project lifecycle — from customer requirements to field deployment in the US and Taiwan.

In 2025 I founded Dakine Robotics to pursue autonomous marine robotics for the recreational boating market.

Technical Skills

Python & QT C++ & ROS Siemens NX / Teamcenter Kinematic Modeling Systems Engineering (V-Model) Embedded Systems Firmware Qualification MATLAB Onshape / Creo LabVIEW Force-Position Control Failure Analysis

Engineering Portfolio

Featured Work

A selection of engineering presentations demonstrating systems thinking, hands-on analysis, and cross-disciplinary problem solving — from semiconductor robotics to mechatronic product design.

Dakine Robotics hull cleaning robot — track-drive deployment
Case Study 01

Dakine Robotics
Autonomous Hull Cleaning Robot

CAD render — isometric view of the hull cleaning robot
Track-drive prototype — workbench view
Electronics architecture schematic
Onboard control electronics
Founder with track-drive prototype on the test vessel
Wheeled prototype — underwater hull attachment test
Video Dakine Robotics test footage

Marine Robotics · Autonomous Systems · Founder-Led R&D

Dakine Robotics — Autonomous Hull Cleaning Robot

Hull biofouling costs the recreational boating industry billions annually in fuel penalties, increased emissions, and environmental fines. What if a robot could live on a hull and keep it permanently clean?

  • Founded Dakine Robotics in 2025 to design and build autonomous underwater hull-cleaning robots for the recreational boating market. Leading all aspects of the venture — mechanical design, electronics, firmware, propulsion, and business development.
  • Built and tested multiple robot prototypes exploring different locomotion strategies: a wheeled propulsion/cleaning variant with a custom 3D-printed propeller and integrated brushroll cleaning assembly, and a rubber track-drive variant optimized for hull surface adhesion and maneuverability.
  • Each prototype features a sealed electronics enclosure, brushless thruster, onboard control electronics, and a modular cleaning head — all designed to operate fully submerged while magnetically or mechanically adhering to a hull surface.
  • Conducting ongoing R&D in marine propulsion, underwater navigation, autonomous path planning, and hull biofouling characterization — with the goal of a robot that deploys autonomously, cleans the full hull, and returns to its dock without user intervention.
  • Developing for the recreational boating market specifically — prioritizing ease of deployment, compact form factor, and low maintenance over industrial-grade complexity.
Marine Robotics Autonomous Systems Propulsion Design 3D Printing Embedded Electronics Hull Biofouling Founder / CEO

Company

Dakine Robotics LLC · Founded Jul 2025

Location

Honolulu, Hawaii

Status

Active R&D — multiple prototypes built and tested

Disciplines

Marine Robotics, Mechanical Design, Embedded Systems, Autonomous Navigation, Propulsion

Purpose

Documenting the problem statement, market research, design, and testing of the first prototype — available on request.

Hybrid Force-Position Control — system test setup
Case Study 02

Hybrid Force-Position Control — MS Capstone

System components — annotated test setup
Force-sensing end effector — mechanical detail
System block diagram
Force-tracking response under ramped command
Steady-state performance data
Video Test setup — robot in working posture

Robotics Research · Graduate Capstone · Santa Clara University

Hybrid Feedforward Force Controller with Selective Position Control Decoupling

How do you make a robot arm that can push against a wall with a precise, controlled force — while simultaneously maintaining exact position on all other axes — and switch seamlessly back to full position control on command?

  • Implemented a hybrid feedforward force controller with selective position control decoupling on a 4-axis Robotis OpenManipulator-X robot arm using ROS and C++. The system can independently control each Cartesian axis in either position or force mode simultaneously — for example, applying a constant X-axis force while maintaining precise Y, Z, and end-effector angle control via PID.
  • The control law combines a PID-based position controller (using the Jacobian inverse and a selection matrix S to isolate position-controlled axes) with a feedforward force controller (using the Jacobian transpose and the complement selection matrix I-S) — summed as τ = τ_pp + τ_ff.
  • Designed, fabricated, and calibrated a custom force-sensing end effector using a 10kg inline load cell, a custom 3D-printed spring compliance housing, and an Arduino/HX711 circuit with LCD display — providing real-time ground-truth force measurement for system validation.
  • Extended the inherited ROS codebase: added free Cartesian coordinate commanding (gotopos), refactored inconsistent kinematic model end effector lengths across four C++ classes (forward kinematics, Jacobian, inertia, gravity), retuned PID and inertia matrix gains to restore stability after the kinematic update, and implemented force ramping (2-second ramp) to eliminate abrupt force transitions.
  • Experimental validation at two arm postures (Z=0.145m, Z=0.185m) across 2N, 4N, and 6N setpoints demonstrated stable force tracking — with best performance at the elevated posture (as low as 2% steady-state error at 4N) — and stable position hold on Y and Z axes throughout force application.
ROS / C++ Hybrid Force-Position Control Jacobian Transpose PID Control Kinematics Custom End Effector Arduino Collaborative Robotics

Purpose

Santa Clara University MS Robotics & Mechatronic Systems capstone thesis — the culminating research project of a seven-year part-time graduate program.

Format

Graduate capstone report + live demo video

Institution

Santa Clara University — MS Robotics & Mechatronic Systems, 2025

Advisor

Dr. Christopher Kitts & Dr. Michael Neumann, SCU Robotic Systems Laboratory

Stack

ROS Noetic, C++, Eigen, Dynamixel SDK, Arduino, Python

Download

65-page capstone report including full source code appendix (C++, Arduino).

Semiconductor Fab
Case Study 03

Automated Semiconductor Robotics Testing Software Suite — LUtil

Factory Interface Robot
Factory Interface Robot in Firmware Test Stand
Applied Materials Opta Platform
Applied Materials Nokota Platform
Semiconductor Tool Automations Layout Diagram
Sample Robot Specifications Sheet
Video Demo Video

Software Engineering · Internal Tooling · Python & Qt

Automated Semiconductor Robotics Testing Software Suite

When manual firmware testing across hundreds of robot speed and motion path permutations became the bottleneck between a software release and the factory floor — how do you fix it once and for everyone?

  • Identified a critical gap in the robotics team's firmware validation process: a "motion path could not be generated" error had reached the field because all permutations of motion paths across all robot speed configurations were never being tested — neither by the robot vendor nor internally.
  • Designed and built LUtil from the ground up in Python with a Qt Designer GUI, enabling any team member to run exhaustive permutation tests across all robot speeds and motion paths with minimal training — prioritizing ease of use and team adoption from day one.
  • Architected LUtil with a modular, expandable test script library so new test cases could be added as the robot product line grew, without requiring changes to the core application framework.
  • Owned the full software development lifecycle: requirements, design, implementation, testing, revision control (Jira), and internal release — the same rigor applied to the robot firmware itself.
  • Also authored and owned robot firmware communication specifications covering robot behaviors, automated teaching sequences, and diagnostic capabilities — making LUtil's test coverage possible end-to-end.
Python Qt / Qt Designer GUI Development Firmware Testing Permutation Testing Jira / Revision Control Semiconductor Robotics

Purpose

A brief overview of the automated test software suite — its architecture, key features, and role in the firmware qualification process.

Format

Internal software tool — documented in semiconductor robotics presentation

Context

Lead developer at Applied Materials, Atmospheric Robotics division, 2017–2025

Stack

Python, Qt, Qt Designer, Jira

Disciplines

Software Engineering, Test Automation, GUI Design, Firmware Communication

Reference

LUtil case study is documented on slides 9–10 of this presentation.

Semiconductor Robotics
Case Study 04

Semiconductor Manufacturing Robotics — Applied Materials

Z-axis Overheating
Job Responsibilities
Motion Path Fix
Inside a Fab

Semiconductor Robotics · Industry Overview & Case Studies

Semiconductor Manufacturing Robotics

What does it actually look like to be the engineer responsible when a robot in a $100M semiconductor fab starts overheating?

  • Explains the semiconductor manufacturing ecosystem from the fab level down to the individual wafer-handling robots inside each tool — context most engineers outside the industry don't have.
  • Case study 1: A fielded z-axis motor overheating intermittently above 90°C at idle. Systematically evaluated five potential fixes, ruled out solutions that would require costly re-qualification, and identified a 10lb mass reduction that doubled the thermal margin — with no re-test required.
  • Case study 2: A "motion path could not be generated" error only reproducible at slow speeds, caused by incomplete permutation testing across all robot speed configurations. Built a Python/QT GUI tool (LUtil) to automate exhaustive permutation testing and enable effortless adoption by the broader team.
  • Covers the full scope of a senior systems engineer's role: system architecture, robot testing, system integration, requirements, supplier management, and field support.
Wafer Handling Root Cause Analysis Python / QT GUI Thermal Analysis Field Engineering Failure Analysis

Format

10-slide PowerPoint presentation

Context

Industry overview and engineering case studies for Elroy Air interview, Mar 2022

Disciplines

Systems Engineering, Thermal Analysis, Python, Semiconductor Manufacturing

Download

⬇ Download Presentation
Kickstand Design
Case Study 05

Automatic Bicycle Kickstand — Systems Design

Motor Sizing
Requirements
Qualification Testing
Design Sketch

Systems Engineering · Interview Assignment

Automatic Bicycle Kickstand

How do you apply aerospace-grade systems engineering discipline to a consumer product — and prove it with math?

  • Decomposed a high-level customer need into a fully traceable requirements hierarchy using the V-model, from stakeholder identification through system requirements and functional breakdown.
  • Developed a Python virtual-work simulation of a four-bar linkage to determine crank torque vs. angle across the full deployment range — resulting in a peak load of 0.144 N·m and a concrete motor selection (3–6 kg·cm DC planetary gear motor).
  • Performed first-order static load calculations by summing moments around the wheel contact point at 6° lean angle to size the structural members.
  • Defined distinct qualification testing (verify the design) vs. pre-deployment acceptance testing (verify each unit) philosophies — a distinction critical to production robotics programs.
  • Documented open design items and assumptions throughout, reflecting real-world engineering rigor rather than a "solved" academic exercise.
V-Model Systems Engineering Python Simulation Four-Bar Linkage Motor Sizing Requirements Decomposition Qualification Testing

Purpose

A sample systems engineering interview assignment completed in 2025 — demonstrating V-model methodology applied to a consumer product from scratch.

Format

20-slide PowerPoint presentation

Context

Take-home systems integration assignment for Reliable Robotics interview, Jan 2025

Disciplines

Systems Engineering, Mechanical Design, Python, Kinematics

Download

Work History

Experience

Dakine Robotics

Jul 2025 – Present

Founder & CEO

Founded and leading an R&D startup developing autonomous hull-cleaning robots for the recreational boating market. Directing research across marine robotics propulsion, hydrodynamic control, and autonomous navigation — solving a costly, largely manual problem in marine maintenance.

Marine RoboticsAutonomous SystemsNavigation & Control

Applied Materials

Apr 2017 – Jan 2025

E3 Senior Robotics Systems Engineer — Atmospheric Robotics

Served as Designated Engineer of Authority (DEA) across multiple business units, owning the full project lifecycle — requirements through design, test, and customer support — for robotic semiconductor wafer-handling systems. Led cross-functional V-model systems engineering with customers and management; managed robot suppliers to performance, cost, and schedule; authored system requirements and interface control documents. Designed high-performance end effectors and motion envelopes in NX/Teamcenter. Built the internal Python/QT LUtil automated firmware testing platform. Executed high-stakes field retrofits and new tool startups at fabs across the US and Taiwan.

Systems EngineeringDEA AuthorityWafer HandlingNX / TeamcenterPython / QTFirmware QualificationField Startups

Applied Materials

May 2015 – Apr 2017

Mechanical Engineering Intern — Atmospheric Robotics

Supported the Atmospheric Robotics team with engineering lab reliability testing, failure analysis, and new product development. Conducted mechanical and software reliability testing with LabVIEW, LCF Sensors, Laser Proximity Sensors, and Vibration Nodes.

LabVIEWReliability TestingFailure Analysis

Space Systems Loral

Jul 2012 – May 2015

Mechanisms & Machine Shop Intern

Assembled satellite flight hardware and supported mechanisms design, test data trending and analysis, and support tooling design. Machine shop roles included precision fabrication, prep and finish, and end-of-line QC inspection.

Satellite HardwareMechanismsQC Inspection

Research & IP

Patents & Publications

Patent

Autoteach Enclosure System (US11370114B2)

Kopec, N.M., Kosinski, L., Farber, M., & Hudgens, J. — United States, Jun 28, 2022

Conference

Design and Control of an Assistive Bionic Joint for Leg Muscle Rehabilitation

Nagel, V., Chu, S., Forney, J., Kosinski, L., & Viswanathan, V. — ASME IMECE 2017, Tampa, FL. Vol. 3: Biomedical and Biotechnology Engineering, V003T04A046

Aptitude

Johnson O'Connor Research Foundation — Exceptional Aptitude Results

Only person in the San Francisco branch's history to score 99th percentile in two aptitudes (Spatial Perception & Structural Visualization) and 95th in a third (Idea Generation).

Academic Background

Education

Santa Clara University

MS — Robotics and Mechatronic Systems (part-time)

Santa Clara, CA · Fall 2017 – Spring 2025

San Jose State University

BS — Mechanical Engineering, Mechatronics Concentration

San Jose, CA · Fall 2013 – Spring 2017

Let's Work Together

Open to opportunities in robotics engineering, mechatronics, and systems engineering. Always happy to talk about marine autonomy or Dakine Robotics.

lylekosinski@gmail.com (650) 793-4430 LinkedIn ↗

Let's Work Together

Lyle Kosinski

Robotics · Mechatronics · Systems Engineering

Email lylekosinski@gmail.com Phone (650) 793-4430 LinkedIn linkedin.com/in/lyle-kosinski

Open to opportunities in robotics engineering, mechatronics, and systems engineering. Always happy to talk about marine autonomy or Dakine Robotics.

⬇ Download