Servoservobrief.com, Saturday, October 31, 2026
A thinner public-market number for Agility anchors the day, with Neura's US-backed Series C and LG's data-and-actuators push filling out the picture.
-
Agility's SPAC lands at $465M, down from the $620M it projected
Agility now expects about $465M gross from its merger with Churchill Capital Corp XI, trading set to start on Nasdaq Nov 5. That is well below the $620M it floated in June at a $2.5B valuation, with $201M of the total a Foxconn-led PIPE (private share purchase) at $10. The shrink shows how much SPAC-trust redemptions and a cooler market can cut a humanoid maker's war chest even as it becomes the first US pure-play humanoid to go public with live deployments.
Read more on Securities.io, The Robot Report, Agility Robotics -
US tech giants back Germany's Neura Robotics in a Series C
US technology companies joined Neura Robotics' latest round to fund its humanoid development, per a founder interview from CEO David Reger. No round size or investor names were disclosed, so the scale and the specific backers remain unconfirmed. It signals US capital reaching into European humanoid makers, not just the domestic and Chinese leaders.
Read more on Funding wire -
LG targets 500,000 hours of robot training data this year, five times its plan
On its Oct 28 earnings call LG said it will bring a 10,000-square-meter data factory in Seoul to full operation this quarter and aim for 500,000 hours of robot training data, up from an original 100,000-hour goal, with a Nvidia Isaac GR00T-based bipedal robot planned for early 2027. Samsung is pushing precision hardware and in-factory testing instead. The two show Korea's split bet: LG on the data and AI layer, Samsung on the hands and actuators.
Read more on BusinessKorea -
Physical Intelligence opens its robot foundation models through an API beta
Physical Intelligence put its latest robot models behind a paid API beta where developers can upload data, fine-tune, and run cloud inference on their own robots. It moves a leading VLA (vision-language-action model, which maps camera input to robot actions) from in-house research toward a product others can build on. No pricing or customer counts disclosed.
Read more on Physical Intelligence