{"552001":{"#nid":"552001","#data":{"type":"news","title":"Tsiotras, Theodorou Help Pioneer New Driverless Car Technology","body":[{"value":"\u003Cp\u003EGT-AE researchers\u003Cstrong\u003E Panagiotis Tsiotras \u003C\/strong\u003Eand \u003Cstrong\u003EEvangelos Theodorou\u003C\/strong\u003E have been workiing\u0026nbsp; with faculty from the School of Interactive Computing (IC) to devise a novel way for the self-driving cars of tomorrow to drive safely under actual road conditions.\u003C\/p\u003E\u003Ctable width=\u0022191\u0022 border=\u00220\u0022 cellspacing=\u00221\u0022 cellpadding=\u00225\u0022 align=\u0022right\u0022\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd\u003E\u0026nbsp;\u003C\/td\u003E\u003C\/tr\u003E\u003Ctr\u003E\u003Ctd class=\u0022rtecenter\u0022\u003E\u0026nbsp;\u003C\/td\u003E\u003C\/tr\u003E\u003C\/tbody\u003E\u003C\/table\u003E\u003Cp\u003EThe team has been quietly testing its work on the Georgia Tech Autonomous Racing Facility on Marietta Street for the last few months, using one-fifth-scale, fully autonomous auto-rally cars that operate at the equivalent of 90 mph. \u003Cbr \/\u003E \u0026nbsp;\u003Cbr \/\u003E Sponsored by the U.S. Army Research Office, the research seeks to increase vehicular stability while maintaining performance. Their technique \u2013 which uses advanced algorithms, onboard computing, and specially installed sensing devices \u2013 is garnering some serious attention, too.\u003C\/p\u003E\u003Cp\u003ELast month, it was presented at the International Conference on Robotics and Automation (ICRA). In the days that followed, it was celebrated in a number of online \u003Ca href=\u0022http:\/\/www.popsci.com\/georgia-tech-builds-an-aggro-autonomous-rally-car\u0022\u003E\u003Cstrong\u003Emedia outlets\u003C\/strong\u003E\u003C\/a\u003E, all of them eager to showcase the next big technological breakthrough in driverless vehicles. \u003Cstrong\u003E\u003Cem\u003E(\u003C\/em\u003E\u003C\/strong\u003E\u003Ca href=\u0022https:\/\/www.youtube.com\/watch?v=1AR2-OHCxsQ\u0022\u003E\u003Cem\u003E\u003Cstrong\u003EThis You-Tube video\u003C\/strong\u003E\u003C\/em\u003E\u003C\/a\u003E\u003Cem\u003E received more than 80K views.) \u003C\/em\u003E\u003C\/p\u003E\u003Ctable width=\u0022160\u0022 border=\u00220\u0022 cellspacing=\u00221\u0022 cellpadding=\u00225\u0022 align=\u0022right\u0022\u003E\u003Ctbody\u003E\u003Ctr\u003E\u003Ctd\u003E\u0026nbsp;\u003C\/td\u003E\u003C\/tr\u003E\u003Ctr\u003E\u003Ctd\u003E\u0026nbsp;\u003C\/td\u003E\u003C\/tr\u003E\u003C\/tbody\u003E\u003C\/table\u003E\u003Cp\u003E\u201cAn autonomous vehicle should be able to handle any condition, not just drive on the highway under normal conditions,\u201d said Tsiotras, an expert on the mathematics behind rally-car racing control. \u003Cbr \/\u003E \u0026nbsp;\u003C\/p\u003E\u003Cp\u003E\u201cOne of our principal goals is to infuse some of the expert techniques of human drivers into the brains of these autonomous vehicles.\u201d\u003C\/p\u003E\u003Cp\u003ETraditional robotic-vehicle techniques use the same control approach whether a vehicle is driving normally or at the edge of roadway adhesion, Tsiotras explained. The Georgia Tech method \u2013 known as model predictive path integral control (MPPI) \u2013 was developed specifically to address the non-linear dynamics involved in controlling a vehicle near its friction limits.\u003Cbr \/\u003E \u0026nbsp;\u003C\/p\u003E\u003Cp\u003E\u003Cstrong\u003EUtilizing Advanced Concepts\u003C\/strong\u003E\u003Cbr \/\u003E \u201cAggressive driving in a robotic vehicle \u2013 maneuvering at the edge \u2013 is a unique control problem involving a highly complex system,\u201d said Evangelos Theodorou, an AE assistant professor who is leading the project. \u201cHowever, by merging statistical physics with control theory, and utilizing leading-edge computation, we can create a new perspective, a new framework, for control of autonomous systems.\u201d\u003C\/p\u003E\u003Cp\u003EThe Georgia Tech researchers used a stochastic trajectory-optimization capability, based on a path-integral approach, to create their MPPI control algorithm, Theodorou explained. Using statistical methods, the team integrated large amounts of handling-related information, together with data on the dynamics of the vehicular system, to compute the most stable trajectories from myriad possibilities.\u003C\/p\u003E\u003Cp\u003EProcessed by the high-power graphics processing unit (GPU) that the vehicle carries, the MPPI control algorithm continuously samples data coming from global positioning system (GPS) hardware, inertial motion sensors, and other sensors. The onboard hardware-software system performs real-time analysis of a vast number of possible trajectories and relays optimal handling decisions to the vehicle moment by moment.\u003C\/p\u003E\u003Cp\u003EIn essence, the MPPI approach combines both the planning and execution of optimized handling decisions into a single highly efficient phase. It\u2019s regarded as the first technology to carry out this computationally demanding task; in the past, optimal- control data inputs could not be processed in real time.\u003C\/p\u003E\u003Cp\u003E\u003Cstrong\u003E Fully Autonomous Vehicles\u003C\/strong\u003E\u003Cbr \/\u003E The researchers\u2019 two auto-rally vehicles \u2013 custom built by the team \u2013 utilize special electric motors to achieve the right balance between weight and power. The cars carry a motherboard with a quad-core processor, a potent GPU, and a battery.\u003Cbr \/\u003E Each vehicle also has two forward-facing cameras, an inertial measurement unit, and a GPS receiver, along with sophisticated wheel-speed sensors. The power, navigation, and computation equipment is housed in a rugged aluminum enclosure able to withstand violent rollovers. Each vehicle weighs about 48 pounds and is about three feet long.\u003C\/p\u003E\u003Cp\u003EThese rolling robots are able to test the team\u2019s control algorithms without any need for off-vehicle devices or computation, except for a nearby GPS receiver. The onboard GPU lets the MPPI algorithm sample more than 2,500, 2.5-second-long trajectories in under 1\/60 of a second.\u003C\/p\u003E\u003Cp\u003EAn important aspect in the team\u2019s autonomous-control approach centers on the concept of \u201ccosts\u201d \u2013 key elements of system functionality. Several cost components must be carefully matched to achieve optimal performance.\u003C\/p\u003E\u003Cp\u003EIn the case of the Georgia Tech vehicles, the costs consist of three main areas: the cost for staying on the track, the cost for achieving a desired velocity, and the cost of the control system.\u003C\/p\u003E\u003Cp\u003EA sideslip-angle cost was also added to improve vehicle stability.\u003Cbr \/\u003E The cost approach is important to enabling a robotic vehicle to maximize speed while staying under control, explained James Rehg, a professor in the Georgia Tech School of Interactive Computing who is collaborating with Theodorou and Tsiotras.\u003C\/p\u003E\u003Cp\u003EIt\u2019s a complex balancing act, Rehg said. For example, when the researchers reduced one cost term to try to prevent vehicle sliding, they found they got increased drifting behavior.\u003C\/p\u003E\u003Cp\u003E\u201cWhat we\u0027re talking about here is using the MPPI algorithm to achieve relative \u003Cbr \/\u003E entropy minimization \u2013 and adjusting costs in the most effective way is a big part of that,\u201d he said. \u201cTo achieve the optimal combination of control and performance in an autonomous vehicle is definitely a non-trivial problem.\u003C\/p\u003E\u003Cp\u003E\u003Cem\u003EStory courtesy of Rick Robinson, Georgia Tech Research News\u003C\/em\u003E\u003C\/p\u003E","summary":null,"format":"limited_html"}],"field_subtitle":[{"value":"Researchers work with School of Interactive Computing"}],"field_summary":[{"value":"\u003Cp\u003EThe collaborative team has been quietly testing its work on the Georgia Tech Autonomous Racing Facility on Marietta Street for the last few months, using one-fifth-scale, fully autonomous auto-rally cars that operate at the equivalent of 90 mph.\u003C\/p\u003E","format":"limited_html"}],"field_summary_sentence":[{"value":"A collaboration with the School of Interactive Computing"}],"uid":"27836","created_gmt":"2016-07-11 15:36:36","changed_gmt":"2016-10-08 03:22:04","author":"Kathleen Moore","boilerplate_text":"","field_publication":"","field_article_url":"","dateline":{"date":"2016-06-06T00:00:00-04:00","iso_date":"2016-06-06T00:00:00-04:00","tz":"America\/New_York"},"extras":[],"hg_media":{"554861":{"id":"554861","type":"image","title":"PanosTsiotras16-9","body":null,"created":"1469382140","gmt_created":"2016-07-24 17:42:20","changed":"1475895353","gmt_changed":"2016-10-08 02:55:53","alt":"PanosTsiotras16-9","file":{"fid":"206576","name":"tsiotras-panagiotis16-9.png","image_path":"\/sites\/default\/files\/images\/tsiotras-panagiotis16-9.png","image_full_path":"http:\/\/www.tlwarc.hg.gatech.edu\/\/sites\/default\/files\/images\/tsiotras-panagiotis16-9.png","mime":"image\/png","size":45443,"path_740":"http:\/\/www.tlwarc.hg.gatech.edu\/sites\/default\/files\/styles\/740xx_scale\/public\/images\/tsiotras-panagiotis16-9.png?itok=UCcQTyqa"}},"501971":{"id":"501971","type":"image","title":"Prof. Evangelos Theodorou","body":null,"created":"1455904800","gmt_created":"2016-02-19 18:00:00","changed":"1475895261","gmt_changed":"2016-10-08 02:54:21","alt":"Prof. Evangelos Theodorou","file":{"fid":"204735","name":"theodorou-evangelos2.jpg","image_path":"\/sites\/default\/files\/images\/theodorou-evangelos2_0.jpg","image_full_path":"http:\/\/www.tlwarc.hg.gatech.edu\/\/sites\/default\/files\/images\/theodorou-evangelos2_0.jpg","mime":"image\/jpeg","size":9643,"path_740":"http:\/\/www.tlwarc.hg.gatech.edu\/sites\/default\/files\/styles\/740xx_scale\/public\/images\/theodorou-evangelos2_0.jpg?itok=_s5rOyGB"}}},"media_ids":["554861","501971"],"groups":[{"id":"1239","name":"School of Aerospace Engineering"}],"categories":[{"id":"136","name":"Aerospace"}],"keywords":[],"core_research_areas":[],"news_room_topics":[],"event_categories":[],"invited_audience":[],"affiliations":[],"classification":[],"areas_of_expertise":[],"news_and_recent_appearances":[],"phone":[],"contact":[],"email":[],"slides":[],"orientation":[],"userdata":""}}}