Search found 119 matches

by johnrossi
Sun Aug 30, 2026 11:59 am
Forum: AI, Control Software & Locomotion
Topic: How close are we to a humanoid recovering from a genuinely novel trip/stumble?
Replies: 16
Views: 358

Re: How close are we to a humanoid recovering from a genuinely novel trip/stumble?

<t>Agreed, and I'd add:

Model predictive control (MPC) is still very much alive in production humanoids, often working alongside or underneath learned policies - MPC handles short-horizon dynamically-consistent trajectory optimization while learned components handle perception, task-level ...
by johnrossi
Sun Aug 30, 2026 11:59 am
Forum: AI, Control Software & Locomotion
Topic: Anyone got good results combining classical footstep planning with a learned recovery policy?
Replies: 26
Views: 309

Re: Anyone got good results combining classical footstep planning with a learned recovery policy?

<t>@rivera14 This is exactly the kind of context I was looking for.

Model predictive control (MPC) is still very much alive in production humanoids, often working alongside or underneath learned policies - MPC handles short-horizon dynamically-consistent trajectory optimization while learned ...
by johnrossi
Sun Aug 30, 2026 11:59 am
Forum: Actuators, Motors & Drivetrains
Topic: How much weight do wiring and connectors actually add per joint?
Replies: 22
Views: 271

Re: How much weight do wiring and connectors actually add per joint?

@servoken70 Worth being a little skeptical of the marketing angle here. Boston Dynamics' commercial electric Atlas switched fully from hydraulic to QDD-style electric actuation, ending up quieter and lighter with comparable or better dynamic performance than the old hydraulic platform.
by johnrossi
Sun Aug 30, 2026 11:59 am
Forum: Actuators, Motors & Drivetrains
Topic: How do you even quantify 'feel' when comparing two actuators?
Replies: 42
Views: 471

Re: How do you even quantify 'feel' when comparing two actuators?

@karen_kim Respectfully, I think this undersells it a bit. Anti-backlash techniques in production actuators range from mechanical preloading (spring-loaded gear meshes) to purely software compensation that models and corrects for known backlash in the control loop.
by johnrossi
Sun Aug 30, 2026 11:59 am
Forum: Manipulation, Grippers & Dexterous Hands
Topic: What's the biggest manipulation failure mode you've personally seen in a demo?
Replies: 49
Views: 775

Re: What's the biggest manipulation failure mode you've personally seen in a demo?

<t>@young58 Counterpoint:

In-hand reorientation (repositioning a grasped object without setting it down) is one of the more advanced manipulation skills, requiring either a highly dexterous hand with enough DoF or clever use of gravity and controlled slipping - it's an active research area rather ...
by johnrossi
Sun Aug 30, 2026 11:59 am
Forum: Manipulation, Grippers & Dexterous Hands
Topic: Anyone using simulation-only training for grasp policies with zero real data?
Replies: 38
Views: 432

Re: Anyone using simulation-only training for grasp policies with zero real data?

<t>@byang Yeah, this tracks with what I've read as well.

Vision-based grasp confidence estimation (predicting success before attempting a grasp) and tactile-based confirmation (confirming after contact) are complementary rather than competing - vision helps you choose a grasp, tactile tells you ...
by johnrossi
Sun Aug 30, 2026 11:59 am
Forum: Manipulation, Grippers & Dexterous Hands
Topic: Anyone tried fully passive compliant fingers with no active sensing at all?
Replies: 45
Views: 535

Re: Anyone tried fully passive compliant fingers with no active sensing at all?

<t>Same conclusion I've come to. Also worth noting:

Cable routing through a wrist joint with multiple degrees of freedom is a genuinely tricky mechanical design problem - tendons and wiring both need enough slack to avoid binding through the full range of motion without tangling or fraying over ...
by johnrossi
Sun Aug 30, 2026 11:59 am
Forum: Power Systems & Thermal Management
Topic: Anyone building their own cell-level monitoring instead of trusting a BMS chip?
Replies: 22
Views: 333

Re: Anyone building their own cell-level monitoring instead of trusting a BMS chip?

<t>Tangent, but worth mentioning:

Hot-swappable battery packs solve the runtime bottleneck for continuous operations (like a 24/7 warehouse shift) without needing a much bigger, heavier pack, but they add mechanical complexity, a failure-prone connector interface, and logistics overhead for ...
by johnrossi
Sun Aug 30, 2026 11:59 am
Forum: DIY & Hobbyist Builds
Topic: What's the cheapest reliable force sensor setup you've actually gotten working?
Replies: 43
Views: 653

Re: What's the cheapest reliable force sensor setup you've actually gotten working?

Same conclusion I've come to. Also worth noting: Joining an existing robotics community (FRC, university clubs, maker spaces) before going fully solo is a common and effective way to pick up practical fabrication and control skills faster than working entirely alone from tutorials.
by johnrossi
Sat Aug 29, 2026 8:11 pm
Forum: DIY & Hobbyist Builds
Topic: Cheapest source for small force/torque sensors for a hobby budget?
Replies: 23
Views: 390

Re: Cheapest source for small force/torque sensors for a hobby budget?

<t>Ran into exactly this myself.

Joining an existing robotics community (FRC, university clubs, maker spaces) before going fully solo is a common and effective way to pick up practical fabrication and control skills faster than working entirely alone from tutorials.

Reminds me a bit of the ...