{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,18]],"date-time":"2026-09-18T06:08:59Z","timestamp":1789711739596,"version":"4.0.1"},"reference-count":28,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,9,18]],"date-time":"2026-09-18T00:00:00Z","timestamp":1789689600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Neurorobot."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Robotic bimanual manipulation poses significant challenges for reactive visuomotor policies, which inherently lack the capacity to anticipate future environmental dynamics, especially in long-horizon and semantically complex tasks. To address this limitation, there is a critical need for frameworks that integrate predictive modeling with task-level reasoning to enable coordinated, forward-looking dual-arm control.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We introduce the World-guided Instruction Semantic Diffusion Policy (WIS), a novel hierarchical framework that tightly integrates a language-driven world model with a diffusion-based policy for dual-arm coordination. The architecture comprises three core modules: (1) a multimodal language\u2013visual perception module that fuses 3D point cloud features with natural language embeddings via cross-modal attention for task-relevant semantic grounding; (2) a language-conditioned world model that explicitly forecasts future state trajectories conditioned on current states, action sequences, and language instructions, providing forward-looking predictive reasoning; and (3) a hierarchical diffusion policy with a high-level latent planner and a low-level diffusion controller, wherein the world model\u2019s predictions actively guide action refinement through a prediction-aware sampling mechanism. This design cleanly separates high-level task reasoning from low-level motor generation while ensuring every action remains grounded in both immediate observations and anticipated future outcomes. We extensively evaluated WIS on the RoboTwin 2.0 platform using Aloha-AgileX bimanual robots across multiple challenging manipulation tasks.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Experimental results on the RoboTwin 2.0 benchmark demonstrate that WIS consistently outperforms strong baselines, achieving superior success rates across all evaluated tasks.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>The proposed method offers an efficient and robust solution for temporally coherent bimanual manipulation, with minimal computational overhead relative to baseline approaches. These findings underscore the value of integrating predictive world models with semantic instructions to enhance both the foresight and dexterity of robotic dual-arm systems in complex, real-world settings.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fnbot.2026.1940438","type":"journal-article","created":{"date-parts":[[2026,9,18]],"date-time":"2026-09-18T05:38:27Z","timestamp":1789709907000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["WIS: world-guided instruction semantic diffusion policy for predictive bimanual manipulation"],"prefix":"10.3389","volume":"20","author":[{"given":"Xukun","family":"Liu","sequence":"first","affiliation":[{"name":"Northwest Institute of Mechanical and Electrical Engineering","place":["Xianyang, Shaanxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Liguo","family":"Hou","sequence":"additional","affiliation":[{"name":"Northwest Institute of Mechanical and Electrical Engineering","place":["Xianyang, Shaanxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Pu","family":"Cheng","sequence":"additional","affiliation":[{"name":"Northwest Institute of Mechanical and Electrical Engineering","place":["Xianyang, Shaanxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiaoping","family":"Luo","sequence":"additional","affiliation":[{"name":"Northwest Institute of Mechanical and Electrical Engineering","place":["Xianyang, Shaanxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fengjuan","family":"Xie","sequence":"additional","affiliation":[{"name":"Northwest Institute of Mechanical and Electrical Engineering","place":["Xianyang, Shaanxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jiqing","family":"Wei","sequence":"additional","affiliation":[{"name":"Northwest Institute of Mechanical and Electrical Engineering","place":["Xianyang, Shaanxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Feng","family":"Wang","sequence":"additional","affiliation":[{"name":"Northwest Institute of Mechanical and Electrical Engineering","place":["Xianyang, Shaanxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aifeng","family":"Liu","sequence":"additional","affiliation":[{"name":"Northwest Institute of Mechanical and Electrical Engineering","place":["Xianyang, Shaanxi, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,9,18]]},"reference":[{"key":"ref1","article-title":"Do as i can, not as i say: grounding language in robotic affordances","author":"Ahn","year":"2022","journal-title":"arXiv"},{"key":"ref2","first-page":"23716","author":"Alayrac","year":"2022"},{"key":"ref3","article-title":"Rt-1: robotics transformer for real-world control at scale","author":"Brohan","year":"2022","journal-title":"arXiv"},{"key":"ref4","article-title":"Robotwin 2.0: a scalable data generator and benchmark with strong domain randomization for robust bimanual robotic manipulation","author":"Chen","year":"2025","journal-title":"arXiv"},{"key":"ref5","doi-asserted-by":"crossref","first-page":"1684","DOI":"10.1177\/02783649241273668","article-title":"Diffusion policy: visuomotor policy learning via action diffusion","volume":"44","author":"Chi","year":"2025","journal-title":"Int. J. Robot. Res."},{"key":"ref6","article-title":"Deep reinforcement learning in a handful of trials using probabilistic dynamics models","volume":"31","author":"Chua","year":"2018","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref7","article-title":"Palm-e: an embodied multimodal language model","author":"Driess","year":"2023","journal-title":"arXiv"},{"key":"ref8","doi-asserted-by":"crossref","first-page":"9156","DOI":"10.52202\/075280-0403","article-title":"Learning universal policies via text-guided video generation","volume":"36","author":"Du","year":"2023","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref9","first-page":"440","article-title":"World models","volume":"2","author":"Ha","year":"2018","journal-title":"arXiv"},{"key":"ref10","article-title":"Dream to control: learning behaviors by latent imagination","author":"Hafner","year":"2019","journal-title":"arXiv"},{"key":"ref11","first-page":"2555","author":"Hafner","year":"2019"},{"key":"ref12","article-title":"Video prediction policy: a generalist robot policy with predictive visual representations","author":"Hu","year":"2024","journal-title":"arXiv"},{"key":"ref13","first-page":"16750","author":"Huang","year":"2023"},{"key":"ref14","article-title":"QT-opt: scalable deep reinforcement learning for vision-based robotic manipulation","author":"Jang","year":"2018","journal-title":"ArXiv"},{"key":"ref15","doi-asserted-by":"publisher","first-page":"394","DOI":"10.3390\/sym18030394","article-title":"PI-VLA: adaptive symmetry-aware decision-making for long-horizon vision\u2013language\u2013action manipulation","volume":"18","author":"Jian","year":"2026","journal-title":"Symmetry"},{"key":"ref16","article-title":"Openvla: an open-source vision-language-action model","author":"Kim","year":"2024","journal-title":"arXiv"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"421","DOI":"10.1177\/0278364917710318","article-title":"Learning hand-eye coordination for robotic grasping with deep learning and large-scale data collection","volume":"37","author":"Levine","year":"2018","journal-title":"Int. J. Robot. Res."},{"key":"ref18","first-page":"19730","author":"Li","year":"2023"},{"key":"ref19","first-page":"29982","author":"Liu","year":"2025"},{"key":"ref20","first-page":"7559","author":"Nagabandi","year":"2018"},{"key":"ref21","first-page":"8748","author":"Radford","year":"2021"},{"key":"ref22","first-page":"77288","author":"Ren","year":"2025"},{"key":"ref23","first-page":"785","author":"Shridhar","year":"2023"},{"key":"ref24","doi-asserted-by":"crossref","first-page":"1344","DOI":"10.3390\/sym17081344","article-title":"FSTGAT: financial spatio-temporal graph attention network for non-stationary financial systems and its application in stock price prediction","volume":"17","author":"Wei","year":"2025","journal-title":"Symmetry"},{"key":"ref25","article-title":"3d diffusion policy: generalizable visuomotor policy learning via simple 3d representations","author":"Ze","year":"2024","journal-title":"arXiv"},{"key":"ref26","article-title":"Socratic models: composing zero-shot multimodal reasoning with language","author":"Zeng","year":"2022","journal-title":"arXiv"},{"key":"ref27","first-page":"2304.13705","article-title":"Learning fine-grained bimanual manipulation with low-cost hardware","author":"Zhao","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref28","first-page":"2165","author":"Zitkovich","year":"2023"}],"container-title":["Frontiers in Neurorobotics"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnbot.2026.1940438\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,18]],"date-time":"2026-09-18T05:38:29Z","timestamp":1789709909000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnbot.2026.1940438\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,18]]},"references-count":28,"alternative-id":["10.3389\/fnbot.2026.1940438"],"URL":"https:\/\/doi.org\/10.3389\/fnbot.2026.1940438","relation":{},"ISSN":["1662-5218"],"issn-type":[{"value":"1662-5218","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,18]]},"article-number":"1940438"}}