Arman Arghavani

Incoming - End Effectors/Hands

Designed 16-DOF tendon-driven robotic hand controlled by 11 actuators achieving 75% human-range grasping with realistic curling kinematics for dextrous manipulation Modeled coupling ratios and joint paths in MATLAB/Simulink; validated mechanics and tendon routing/behavior in MuJoCo Optimized embedded tendon paths by reducing curves and direction changes to reduce friction losses and improve actuator-to-joint power transfer Conducted root cause analysis of finger curling singularities caused by MCP-driven tendon length changes in PIP and DIP joints, guiding actuator and control strategy selection to preserve joint position and force capability Designed compact forearm-mounted actuator packaging instead of palm-mounted to reduce counter torque on elbow joint from motors by 48%

Developing a biomimetic tendon-driven humanoid hand. • Designed & prototyped a 15-DOF tendon-driven hand with natural joint coupling • Built tendon-routing system improving force transmission & reducing friction • Servo-actuated DOFs and validated finger kinematics through physical testing • Co-authored research on hybrid underactuated tendon control vs fully actuated systems

Autonomous racing platform (F1-Tenth). • Developed reactive collision-avoidance in C++ & ROS2 with real-time control tuning • Improved Follow-the-Gap by adding race-line prediction for faster trajectories • Designed & FEA-optimized LiDAR rack; reduced weight ~15% while improving rigidity • Fabricated mounts & welded frame with vibration-damping features

Designing robot end-effectors on PAL Robotics’ TIAGo Pro humanoid for precise, compliant grasping in chemical lab settings; integrated force/tactile sensors and depth cameras for closed-loop control Developed directionally compliant TPU gripper for soft manipulation, allowing 50% compression under maximum operating load while resisting shear-induced deformation Validated gripper compression in ANSYS using nonlinear hyperelastic FEA, confirming predictable force–displacement and shape retention Created 2-joint locking camera mount with 180° articulation for quick angle changes without redesign/reprinting, reducing iteration time by 10 hrs

Building Humanoids Backed by Lux Capital & Refactor Capital

Built autonomous food-preparation and cooking robots. Founder’s Inc F25 DraperU pitch competition 3rd/70 teams S25

• Supported testing & validation of Communication-Based Train Control (CBTC) automation systems • Drew and interpreted SCADA and electrical schematics for train control + rail infrastructure systems • Performed thermal load analysis & lifecycle costing for server-room HVAC and mission-critical components • Assisted reliability & redundancy assessments for passenger rail safety systems

Contracted to work on the Communications Based Train Control project at Bay Area Rapid Transit.

Designing 3D Models and animating them to be used in-game. Roblox game “Jujutsu Kaisen: King of Curses” has gained 70,000+ plays. 3D modeling and animating parts using Blender and Roblox Studio. Integrating game components with LUA programming language.

• Led a team of fifteen students to prototype and build a wind turbine that produced 200J with 40% efficiency. • Placed 1st in California 2x (2023 and 2024), and 6th and 7th in the US (2023 and 2024). • Applied CFD analysis to refine blade design, increasing power output by 20% through iterative optimization. • Partnered with UCLA and Cal Maritime to test blades at various wind speeds, iterating airfoils for 30% efficiency boost.

Researching innovative ways to improve the field of medicine. 2nd place in College Physics at the 2023 HOSA State Leadership Conference. 6th place in College Physics at the 2024 HOSA State Leadership Conference. 7th place in Job Seeking Skills at the 2024 HOSA State Leadership Conference. Invited to compete at the 2023 HOSA International Leadership Conference.

President’s Scholarship Term Distinction