The Robotics Podcast with Fexingo: Autonomous Systems, Industrial Robots, and Hardware

In this episode of The Robotics Podcast, Lucas and Luna explore a deceptively hard challenge for robots: opening a door. They break down the physics of latching mechanisms, the force-torque sensing problem, and why a task humans find trivial remains a frontier in manipulation. The episode contrasts industrial robots that can lift a car door with humanoids that still fumble a push-bar, drawing on r...

Lucas and Luna dive into one of robotics' most stubborn problems: grasping deformable objects. They use the specific case of a tennis ball — something a human hand can grab without thinking — to illustrate why even the most advanced robot arms fail at it. They break down the physics of compliance, the limits of current force-torque sensors, and why deep learning alone hasn't solved the problem. Th...

In this episode of The Robotics Podcast, Lucas and Luna explore why even the most advanced robotic arms struggle to handle high-variation parts like car doors. They discuss the core problem of tolerance stacking in automotive assembly, how robots handle stiff parts, and why compliant parts cause chaos. The episode highlights real-world examples from automotive factories and why human workers remai...

Episode 26 of The Robotics Podcast dives into an everyday frustration that exposes a deep gap in robotics: inserting a USB plug correctly. Lucas and Luna explore why humans do it effortlessly while robots fail 30 percent of the time on the first try. They break down the geometry problem, the role of tactile sensing, and a 2025 paper from MIT CSAIL that used force-controlled search strategies to re...

In Episode 25 of The Robotics Podcast, Lucas and Luna tackle one of robotics' most stubborn unsolved problems: cable manipulation. Why can't a robot arm reliably plug in a USB cable or route a charging cord? The hosts dive into the physics of flexible objects — how cables have near-infinite degrees of freedom, why force-torque sensors struggle with variable stiffness, and why even deep learning fa...

Knot-tying is one of the hardest manipulation tasks for robots — it requires precise force control, complex trajectory planning, and handling deformable objects. In this episode, Lucas and Luna explore why knot-tying remains an open research problem despite decades of robotics progress. They discuss how the DARPA hackathon approach pushed teams to failure, why tying shoelaces is harder than surgic...

Episode 23 of The Robotics Podcast digs into a hidden bottleneck in industrial robotics: force-torque sensor reliability. Lucas explains why even a $100,000 robot arm can't tell if it's tightening a bolt or crushing an egg, and Luna brings in a real-world case from a battery assembly line where false sensor readings caused a 12% scrap rate. They discuss the physics behind sensor drift, why calibra...

Episode 22 of The Robotics Podcast picks a deceptively simple task: tightening a screw. Lucas explains why a torque-controlled robot arm still strips threads or fails to seat bolts, using the production lines at a German automotive supplier as a case study. Luna asks whether recent advances in force-torque sensors and reinforcement learning have closed the gap. The hosts get into the physics of fr...

Episode 21 of The Robotics Podcast with Fexingo dives into a deceptively hard problem: tightening a bottle cap. Lucas and Luna break down why torque control, compliance, and tactile sensing make this simple human gesture a nightmare for industrial robot arms. They explore a 2025 paper from MIT's CSAIL that benchmarks six robot arms on cap-tightening, revealing that even the best models fail 40% of...

Lucas and Luna dig into a surprisingly hard robotics problem: packing a suitcase. Unlike folding a towel or picking an apple, suitcase packing requires reasoning about irregular 3D space, varying object stiffness, and dynamic constraint satisfaction. Lucas walks through how even the most advanced robot arms from Amazon's Sparrow system and Boston Dynamics' Stretch struggle with arbitrary container...

Lucas and Luna dig into one of robotics' most stubborn problems: dextrous manipulation of soft, irregular objects. They focus on why peeling a potato is surprisingly harder than assembling a car, and how a team at Carnegie Mellon used a $200 force sensor and a lot of machine learning to get closer. The episode covers the physics of variable stiffness, the limits of current grippers, and why a simp...

Lucas and Luna explore why robot sensors struggle in high-humidity environments like food processing plants and coastal warehouses. They focus on a 2025 incident at a seafood processor where condensation caused a vision-guided gripper to drop 12 percent of its catch. The hosts break down the physics of fogging and condensation on LiDAR and cameras, discuss current mitigation strategies like hydrop...

Lucas and Luna explore the hidden bottleneck in industrial robotics: the physical move. When a car maker relocates a robotic assembly line, the process can take weeks—even months—because each robot's kinematic model, safety zone, and tool-center-point calibration must be rebuilt from scratch. Lucas unpacks how one German automotive supplier, ZF Friedrichshafen, spent 47 days moving 112 robots from...

Lucas and Luna dive into why lithium-ion battery fires are a growing concern in robotics, from warehouse fleets to humanoid prototypes. Using the 2025 recall of 40,000 Locus Robotics bots as a case study, they explain the chemistry, the thermal runaway risk, and what engineers are doing about it. They also touch on how battery safety could shape the next generation of autonomous hardware and why f...

Lucas and Luna break down why humanoid robots — despite all the splashy demos from Tesla, Figure, and Agility Robotics — are still a long way from replacing human workers. They focus on the specific bottleneck of energy density: today's best batteries can power a humanoid for only about an hour of continuous work, compared to a human's 8-hour shift. Lucas brings numbers from a 2025 Boston Dynamics...

In episode 14 of The Robotics Podcast, Lucas and Luna explore a surprising frontier in robotics: why even the most advanced robot arms struggle with tasks that humans find trivial, like threading a needle or handling a raw egg without breaking it. They dive into the physics of contact-rich manipulation, the limitations of current tactile sensors, and the latest research from MIT's CSAIL lab on 'ge...

Lucas and Luna dive into one of robotics' most deceptive challenges: deformable object manipulation. Why can a robot assemble a car engine but still fail to fold a towel or handle a bedsheet? The hosts break down the physics problem of infinite degrees of freedom in fabric, the limitations of current grippers and vision systems, and what breakthroughs like those from UC Berkeley's AUTOLab and rese...

Episode 12 of The Robotics Podcast dives into robot maintenance and downtime — the hidden cost of automation that no sales brochure mentions. Lucas and Luna discuss the real-world economics of keeping industrial robots running, using the 2023 Ford Kansas City plant shutdown as a case study. They break down why a single failed sensor can cost $100,000 an hour, the tension between lean manufacturing...

Episode 11 explores why welding, a 3,000-year-old process, remains one of the most difficult tasks to automate in manufacturing. Lucas and Luna examine the physics, sensor challenges, and economics behind the 'weld gap' problem. Using the example of a small automotive supplier in Ohio that spent $2 million trying to automate a single welding cell — and still kept one human operator on site — they...

Episode 10 dives into the surprising regulatory paradox facing autonomous taxi fleets. Lucas and Luna examine why robotaxis in San Francisco and Phoenix are being held back not by technology failures but by the very safety data they're required to report. The hosts break down how a 2024 California DMV report showed autonomous vehicles had a 75% lower collision rate per mile than human drivers, yet...

Lucas and Luna examine why multi-robot coordination remains one of the hardest unsolved problems in robotics. They break down the specific failure modes — from communication bandwidth limits in dense swarms to the computational cost of collision-free path planning for just 20 robots. Lucas contrasts the brute-force approach taken by Amazon's warehouse fleet (centralized orchestration) with the dec...

Episode 8 of The Robotics Podcast with Fexingo: Lucas and Luna explore the uncomfortable truth about robot energy efficiency. They compare the power draw of a single industrial robot arm to a solar panel array, unpack why energy consumption remains robotics' silent bottleneck, and discuss how a new wave of low-power actuators is changing the math. With specific numbers from Tesla's factory footpri...

Lucas and Luna dive into the hidden bottleneck slowing down industrial robotics: communication protocols. While robot hardware gets faster every year, the software standards that let robots talk to each other—and to factory management systems—are stuck in the 1990s. Lucas breaks down the specific problem of 'interoperability latency,' using the real-world example of a mid-size German auto parts su...

Episode 6 of The Robotics Podcast tackles a deceptively hard problem: why industrial cameras still get blinded by glare, even on million-dollar assembly lines. Lucas and Luna dig into a 2025 case from a German automotive tier-1 supplier that lost an entire shift because a robot mistook a reflection for a part. They unpack the physics — dynamic range limits, polarized lighting retrofits, and the ga...

Episode 5 of The Robotics Podcast with Fexingo dives into the quiet revolution in end-effector design—soft grippers. Lucas and Luna examine why traditional rigid claws fail on delicate objects like fruit or glass, and how startups like Soft Robotics Inc. are using pneumatic bladders and gecko-inspired adhesives to handle everything from eggs to engine parts. The hosts break down the engineering tr...