[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84140-en":3,"doc-seo-84140-105":29,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":11,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},84140,687197207057,"Sage","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","Embodied Human-Robot Interaction via Acoustics A MARL Approach with AcoustoBots for Spatial Data Physicalization","Traditional data physicalization is often static and disconnected from real environments, weakening its ability to convey embodied spatial dynamics and support rapid user sense-making in situ. AcoustoBots provides a mobile acoustophoretic physicalization platform: TurtleBot3 robots use upward-facing 8×8 ultrasonic phased arrays to levitate particles whose height (1–10 cm) encodes local urban scalar values such as population density, noise, and traffic. A centralized-training, decentralized-execution MARL policy (MADDPG) selects collision-aware navigation actions while a high-rate GS-PAT acoustic controller maintains trap stability and enforces the commanded height, forming a closed perception–display–action loop. Single- and dual-robot evaluations on scaled maps show stable levitation and consistent location-dependent height rendering, with 90% and 80% task success and low collisions across repeated trials.","Embodied Human–Robot Interaction via Acoustics: A MARL Approach with AcoustoBots for Spatial Data Physicalization  \nShiqi Liu  \nDepartment of Computer Science University College London London, United Kingdom [zczqs66@ucl.ac.uk](zczqs66@ucl.ac.uk)  \nNarsimlu Kemsaram  \nDepartment of Artificial Intelligence University of Malaya Kuala Lumpur, Malaysia[narsimlu.kemsaram@um.edu.my](narsimlu.kemsaram@um.edu.my)  \nPrateek Mittal  \nDepartment of Computer Science University College London London, United Kingdom [prateek.mittal@ucl.ac.uk](prateek.mittal@ucl.ac.uk)  \nPengyuan Wei  \nDepartment of Computer Science University College London London, United Kingdom [pengyuan.wei.22@ucl.ac.uk](pengyuan.wei.22@ucl.ac.uk)  \nSriram Subramanian  \nDepartment of Computer Science University College London London, United Kingdom[s.subramanian@ucl.ac.uk](s.subramanian@ucl.ac.uk)  \narXiv :2607 .06563v 1 [ cs .RO] 7 Jul 2026  \nFigure 1: Overview of the proposed AcoustoBots platform for embodied spatial data physicalization and robot-mediated communication. AcoustoBots navigate on a projected UK grid map and sample local urban scalar data, such as population density, noise, or traffic. Each TurtleBot3 carries an upward-facing 8×8 ultrasonic phased array that levitates a particle whose height (1–10 cm) encodes the local scalar value as a glanceable physical cue. A MARL policy selects safe and informative navigation actions from robot state and physicalization-related observations, while a high-rate GS-PAT-based acoustic controller continuously updates array phases to maintain stable levitation and control the commanded particle height during motion, thereby closing the perception–display–action loop.  \nAbstract  \nTraditional data physicalization is often static and disconnected from real environments, limiting its ability to convey embodied spatial dynamics and engage users. To address this limitation, we present AcoustoBots, a mobile acoustophoretic data-physicalization platform in which TurtleBot3 robots carry upward-facing 8 × 8 ultrasonic phased arrays. Each array levitates a particle whose height (1–10 cm) encodes a local urban scalar value, such as population density, noise, or traffic. A MARL (Multi-Agent Reinforcement Learning) policy based on the Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm, with centralized training and decentralized execution, selects collision-aware navigation actions, while a high-rate Gerchberg-Saxton-Phased Array of Transducers (GS-PAT) acoustic controller maintains trap stability and updates array phases to achieve the commanded height during motion. This creates a closed perception–display–action loop. We evaluate singlerobot city-to-city traversal and dual-robot cooperative coverage on a 4 m x 3 m scaled UK map using PhaseSpace-based localization  \nfor repeatable multi-robot trials. Results show stable in-motion levitation and consistent, location-dependent height rendering, with task success rates of 90% and 80% for the single-and dual-robot regimes, respectively, over 10 trials per regime, and low collision counts. These findings support acoustophoretic levitation as a simple, glanceable, robot-mediated communication cue for embodied human–robot interaction in spatial analytics.  \nCCS Concepts  \n• Human-robot interaction → Interaction techniques; • Computer systems organization → Robotic autonomy; External interfaces for robotics; • Computing methodologies → Reinforcement learning; Spatial data physicalization.  \nKeywords  \nAcoustoBots, multi-agent reinforcement learning, acoustophoretic interaction, contactless object manipulation, human-robot interaction, swarm robotics.  \n1 INTRODUCTION  \nCities increasingly depend on timely, spatially resolved information—such as noise exposure, traffic flow, air quality, and population dynamics—to guide planning and operations [1] . Yet the dominant paradigm still routes sensing through offline pipelines and then visualizes results on flat screens that are spat","cbCaiiIAKgwKkkuM","https://ap.wps.com/l/cbCaiiIAKgwKkkuM","pdf",21732641,2,1,"English","en",105,"# 1 Introduction\n# CCS Concepts\n# Keywords","[{\"question\":\"What problem does AcoustoBots target in data physicalization?\",\"answer\":\"AcoustoBots addresses the limitation of traditional physicalization being static and spatially detached from the real environment, which hinders embodied understanding and quick sense-making in situ.\"},{\"question\":\"How does AcoustoBots encode spatial data for users?\",\"answer\":\"Each TurtleBot3 carries an upward-facing 8×8 ultrasonic phased array that levitates a particle; the particle height (1–10 cm) encodes local urban scalar values like population density, noise, or traffic.\"},{\"question\":\"How are navigation and acoustic control coordinated in the system?\",\"answer\":\"A MARL policy selects collision-aware navigation actions using observations related to physicalization, while a high-rate GS-PAT acoustic controller updates array phases to maintain trap stability and realize the commanded particle height during motion.\"}]",1784193342,20,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":27},"embodied-human-robot-interaction-via-acoustics-a-marl-approach-with-acoustobots-for-spatial-data-physicalization","",{"@graph":35,"@context":84},[36,52,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,46,49],{"item":40,"name":41,"@type":42,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":20},"https://docshare.wps.com/document/","Document",{"item":47,"name":12,"@type":42,"position":48},"https://docshare.wps.com/document/research-report/",3,{"item":50,"name":13,"@type":42,"position":51},"https://docshare.wps.com/document/embodied-human-robot-interaction-via-acoustics-a-marl-approach-with-acoustobots-for-spatial-data-physicalization/84140/",4,{"url":50,"name":13,"@type":53,"author":54,"headline":13,"publisher":56,"fileFormat":59,"inLanguage":23,"description":14,"dateModified":60,"datePublished":61,"encodingFormat":59,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":55},"Person",{"url":40,"name":57,"@type":58},"DocShare","Organization","application/pdf","2026-07-25","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":20},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What problem does AcoustoBots target in data physicalization?","Question",{"text":74,"@type":75},"AcoustoBots addresses the limitation of traditional physicalization being static and spatially detached from the real environment, which hinders embodied understanding and quick sense-making in situ.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"How does AcoustoBots encode spatial data for users?",{"text":79,"@type":75},"Each TurtleBot3 carries an upward-facing 8×8 ultrasonic phased array that levitates a particle; the particle height (1–10 cm) encodes local urban scalar values like population density, noise, or traffic.",{"name":81,"@type":72,"acceptedAnswer":82},"How are navigation and acoustic control coordinated in the system?",{"text":83,"@type":75},"A MARL policy selects collision-aware navigation actions using observations related to physicalization, while a high-rate GS-PAT acoustic controller updates array phases to maintain trap stability and realize the commanded particle height during motion.","https://schema.org",{"og:url":50,"og:type":86,"og:title":13,"og:site_name":57,"og:description":14},"article",{"robots":88,"canonical":50},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,122,126,129,133],{"id":21,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":20,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":51,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":28,"slug":125},9,"Religion & Spirituality","religion-spirituality",{"id":28,"doc_module":4,"doc_module_name":45,"category_name":127,"show_sort_weight":28,"slug":128},"World Cup","world-cup",{"id":130,"doc_module":4,"doc_module_name":45,"category_name":131,"show_sort_weight":130,"slug":132},10,"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":45,"category_name":135,"show_sort_weight":105,"slug":136},19,"General","general"]