[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84166-en":3,"doc-seo-84166-105":28,"detail-sidebar-cat-0-en-105":89},{"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":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":11,"language":21,"language_code":22,"site_id":23,"html_lang":22,"table_of_contents":24,"faqs":25,"seo_title":13,"seo_description":14,"update_tm":26,"read_time":27},84166,1374391974468,"Eden","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","Ace Motion Planning of Professional-Level Table Tennis Serves with a Robot Arm","Table tennis serves become an important yet underexplored robotics benchmark because competitive play requires extreme physics modeling and precise control beyond rally interception. A method generates International Table Tennis Federation–compliant serves by combining motion primitives, Model Predictive Control, and Bayesian Optimization. The pipeline produces controllable spin up to 550 rad/s and serve speeds up to 6.7 m/s, matching and exceeding elite human performance while enabling accurate aiming and multi-objective optimization.","Ace! Motion Planning of Professional-Level Table Tennis Serves with a Robot Arm  \nGuillem Torrente 1 , Guilherme Jorge Maeda 1 , Divij Grover2 , Megumu Tsukamoto3 , Hamdi Sahloul 1   \nand Peter Dürr2 , Senior Member, IEEE  \narXiv :2607 .06989v 1 [ cs .RO] 8 Jul 2026  \nAbstract—Table tennis, a dynamic, compact, and popular sport, has received significant attention as a robotics benchmark over the last decades. Most of the research has focused on the rally aspect - returning an incoming ball - requiring highspeed vision, agile motion planning, and tight closed-loop control. However, the other component of table tennis gameplay - the serve - is comparatively a quite unexplored research problem, that in fact requires pushing physics modeling and control to the extremes. Achieving competitive serves with a robot presents domain-specific challenges, such as high-spin generation from aspinless ball, precise aiming, or multi-objective optimization. In this work, we present a novel approach for generating official rule-compliant serves by combining motion primitives, Model Predictive Control, and Bayesian Optimization. Serves generated in this way offer a wide and controllable variation of spins of up to 550 rad/s, and speeds of up to 6.7 m/s, matching and even surpassing those of elite table tennis players.  \nIndex Terms—Model Predictive Control, dynamic object interception, HEBO, motion planning, robot arm.  \nI. INTRODUCTION  \nTable tennis is widely regarded as a rigorous benchmark for high-speed robotics. Characterized by its compact playing field, rapid ball exchanges, and complex aerodynamics, the sport demands systems capable of exceptional visual acuity, low-latency decision-making, and agile actuation [1, 2] . Over the past several decades, the robotics research community has extensively utilized table tennis to advance the state of the art in these fields [3, 4] .  \nHistorically, the vast majority of research in this domain has focused on the \"rally\", that is, the problem of intercepting and returning an incoming ball [5, 6] . On the other hand, serving with a robot has received comparatively much less attention as a research problem. This focus is understandable; the challenges of the rally map directly to fundamental and universal requirements in robotics, which necessitate robust solutions for dynamic environment perception and reactive motion planning. Consequently, significant progress has been made in developing robot systems capable of sustaining rallies against human opponents [7–9] .  \n1 G. Torrente, G. J. Maeda, and H. Sahloul are with Sony AI, Tokyo, Japan.  \n2D. Grover and P. Dürr are with Sony AI, Zürich, Switzerland.  \n3M. Tsukamoto is with Sony Group Corporation, Tokyo, Japan.  \nHuman table tennis players competed against our robot for the purpose of performance evaluation. Players were not subject to any situation involving risk. There was no direct physical contact between the two parties. No private data was collected other than videos of the experiments, with prior informed consent. Experiments involved competitive matches between humans and the robot according to official game settings, including umpires.  \nFig. 1. A high side-spin serve executed by the robot. The ball trajectory curves after each bounce due to the spin applied. The robot pose corresponds to approximately that at the moment of contact with the ball.  \nThe serve aspect remains relatively underexplored as far as robotics research is concerned, despite being an important part of the table tennis game. Being the only fully controllable action during the point, elite players use the serve tactically to force certain actions on the opponent and try to secure a quick advantage [10] . Unlike a return, where the player leverages the momentum on the incoming ball, a competitive serve requires the generation of high-speed and/or high-spin from a spinless, free-falling ball [2, 11] . But, since the ball must bounce first on the player’s own side of the","cbCaiqbJJDyxC5ZS","https://ap.wps.com/l/cbCaiqbJJDyxC5ZS","pdf",3830728,1,"English","en",105,"# Introduction\n## Serve as an Underexplored Robotics Problem\n## Contributions and System Pipeline","[{\"question\":\"Why is table tennis serving harder for robots than rally returns?\",\"answer\":\"Serving requires generating high-speed and/or high-spin from a spinless, free-falling ball, while also enforcing bounce constraints and very precise interception angles.\"},{\"question\":\"What core techniques are used to generate competitive robot serves?\",\"answer\":\"The approach combines motion primitives with Model Predictive Control and uses Bayesian Optimization (via HEBO) to optimize motion-planner parameters such as hitting state and timing.\"},{\"question\":\"What performance ranges do the generated serves achieve?\",\"answer\":\"Generated serves reach spin up to 550 rad/s and speeds up to 6.7 m/s, with controllable variation that matches and can surpass elite table tennis players.\"}]",1784193596,20,{"code":4,"msg":29,"data":30},"ok",{"site_id":23,"language":22,"slug":31,"title":13,"keywords":32,"description":14,"schema_data":33,"social_meta":84,"head_meta":86,"extra_data":88,"updated_unix":26},"ace-motion-planning-of-professional-level-table-tennis-serves-with-a-robot-arm","",{"@graph":34,"@context":83},[35,52,66],{"@type":36,"itemListElement":37},"BreadcrumbList",[38,42,46,49],{"item":39,"name":40,"@type":41,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":43,"name":44,"@type":41,"position":45},"https://docshare.wps.com/document/","Document",2,{"item":47,"name":12,"@type":41,"position":48},"https://docshare.wps.com/document/research-report/",3,{"item":50,"name":13,"@type":41,"position":51},"https://docshare.wps.com/document/ace-motion-planning-of-professional-level-table-tennis-serves-with-a-robot-arm/84166/",4,{"url":50,"name":13,"@type":53,"author":54,"headline":13,"publisher":56,"fileFormat":59,"inLanguage":22,"description":14,"dateModified":60,"datePublished":60,"encodingFormat":59,"isAccessibleForFree":61,"interactionStatistic":62},"DigitalDocument",{"name":9,"@type":55},"Person",{"url":39,"name":57,"@type":58},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":63,"interactionType":64,"userInteractionCount":4},"InteractionCounter",{"@type":65},"ViewAction",{"@type":67,"mainEntity":68},"FAQPage",[69,75,79],{"name":70,"@type":71,"acceptedAnswer":72},"Why is table tennis serving harder for robots than rally returns?","Question",{"text":73,"@type":74},"Serving requires generating high-speed and/or high-spin from a spinless, free-falling ball, while also enforcing bounce constraints and very precise interception angles.","Answer",{"name":76,"@type":71,"acceptedAnswer":77},"What core techniques are used to generate competitive robot serves?",{"text":78,"@type":74},"The approach combines motion primitives with Model Predictive Control and uses Bayesian Optimization (via HEBO) to optimize motion-planner parameters such as hitting state and timing.",{"name":80,"@type":71,"acceptedAnswer":81},"What performance ranges do the generated serves achieve?",{"text":82,"@type":74},"Generated serves reach spin up to 550 rad/s and speeds up to 6.7 m/s, with controllable variation that matches and can surpass elite table tennis players.","https://schema.org",{"og:url":50,"og:type":85,"og:title":13,"og:site_name":57,"og:description":14},"article",{"robots":87,"canonical":50},"index,follow",{"doc_id":7,"site_id":23},{"code":4,"msg":5,"data":90},[91,95,99,103,108,113,118,121,125,128,132],{"id":20,"doc_module":4,"doc_module_name":44,"category_name":92,"show_sort_weight":93,"slug":94},"Story & Novel",90,"story-novel",{"id":45,"doc_module":4,"doc_module_name":44,"category_name":96,"show_sort_weight":97,"slug":98},"Literature",80,"literature",{"id":51,"doc_module":4,"doc_module_name":44,"category_name":100,"show_sort_weight":101,"slug":102},"Exam",70,"exam",{"id":104,"doc_module":4,"doc_module_name":44,"category_name":105,"show_sort_weight":106,"slug":107},5,"Comic",60,"comic",{"id":109,"doc_module":4,"doc_module_name":44,"category_name":110,"show_sort_weight":111,"slug":112},6,"Technology",50,"technology",{"id":114,"doc_module":4,"doc_module_name":44,"category_name":115,"show_sort_weight":116,"slug":117},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":44,"category_name":12,"show_sort_weight":119,"slug":120},30,"research-report",{"id":122,"doc_module":4,"doc_module_name":44,"category_name":123,"show_sort_weight":27,"slug":124},9,"Religion & Spirituality","religion-spirituality",{"id":27,"doc_module":4,"doc_module_name":44,"category_name":126,"show_sort_weight":27,"slug":127},"World Cup","world-cup",{"id":129,"doc_module":4,"doc_module_name":44,"category_name":130,"show_sort_weight":129,"slug":131},10,"Lifestyle","lifestyle",{"id":133,"doc_module":4,"doc_module_name":44,"category_name":134,"show_sort_weight":104,"slug":135},19,"General","general"]