−8.9%
since Oct 8
The idea
I believe consumer robotics is about to move from demos to real shipping volume, and every humanoid, home robot, or robotaxi needs the same parts: precise motors and bearings, sensors that see, chips that think, and power electronics to run it all. Rather than bet on which robot maker wins, I own the suppliers already landing design wins across motion, perception, compute, and power, spread across many names so no single stock carries the thesis. I keep 10% in cash for dips. If design win announcements dry up for a couple of quarters, I consider the idea broken, not just slow.
What it owns
- The take needs motors, gearing, bearings, and actuators to see rising orders from robot and automation programs. · Regal RexnordIt makes the motors, gearing, and power transmission parts that robots depend on to move, and needs robotics to grow as a share of its orders., NovantaIt sells precision motion and photonics components into robotics and automation, and needs to keep winning design sockets in new robot programs., TimkenIt makes engineered bearings that sit in every joint of robot arms and humanoids, and needs robotics to become a growing slice of its orders., Parker HannifinIt dominates motion control systems, actuators, and hydraulics that scale from factory robots to humanoids, and needs automation demand to keep growing., RBC BearingsIt makes highly engineered bearings for demanding uses in humanoid joints and robotaxi drivetrains, and needs precision bearing demand to broaden from aerospace into robotics., AllientIt makes small precision motors and motion controllers that humanoid hands and home robots need in quantity, and needs robotics orders to grow visibly.28%
- The take bets that machine vision, lidar, imaging, and motion sensing become critical and high-volume parts of every deployed robot. · CognexIt is the standard in machine vision cameras and software that robots use to see and inspect, and holds its lead as robot volumes grow., onsemiIt supplies image sensors and power chips for robots and vehicles that navigate the real world, and needs to win sockets in robotaxi and robot platforms., TE ConnectivityIt makes the connectors and sensors that integrate every robot's parts together, with high content per machine as volumes rise., TeledyneIt supplies industrial imaging, infrared, and specialty sensors that give robots sight beyond ordinary cameras, and needs machine vision demand to keep growing., STMicroelectronicsIt makes motion sensors, microcontrollers, and motor-driver chips for consumer and industrial robots, and needs its robotics lines to win sockets in the wave of new robots., OusterIt sells digital lidar sensors that robots and robotaxis use to map their surroundings, and needs lidar to stay essential for autonomy and profitability.25%
- The take bets on standardization around a few chip platforms and software stacks that train and run most robot programs. · NVIDIAIt supplies the compute and simulation stack that most humanoid and robotaxi programs standardize on for training and deployment., QualcommIt powers battery-constrained robots like home robots and drones through its low-power chip designs, and needs its robotics platform to keep winning beyond phones., Arm HoldingsIt licenses the chip designs that power most low-power robot brains, earning a royalty on nearly every deployed robot regardless of chipmaker.18%
- The take needs power management and conversion chips to see growing demand as robots pack more compute and motors into tight spaces. · Monolithic Power SystemsIt makes the power management chips that feed compute and motors efficiently, and needs its automation and compute power business to outpace the broader chip market., Texas InstrumentsIt supplies the analog and embedded chips that sit between a robot's brain and its motors and sensors, with dozens per machine., VicorIt makes compact high-density power modules suited to power-hungry robot compute, and needs robotics and AI design wins to convert into shipping revenue.12%
- The take bets directly on two companies proving that robot programs can move from pilots to production and paying customers. · TeslaIt is one direct bet on the robotics inflection through Optimus and robotaxi programs, and needs visible progress from pilots toward production and paying customers., SymboticIt is the second direct inflection bet with warehouse robots already deployed and generating revenue, and needs its backlog to convert and new customers to sign beyond anchor deals.7%
- Cash10%