BEGIN:VCALENDAR
PRODID:-//vBulletin 6//EN
VERSION:2.0
CALSCALE:GREGORIAN
BEGIN:VEVENT
UID:9301eeb0-3783-4625-90ff-398c25aaf1fd
DTSTAMP:20260519T134359Z
SUMMARY:AutoDriving CTF\, Friday 10:00 - Saturday 18:00
DESCRIPTION:.. [/URL]\n \n\n\n\n..\n\nAutoDriving CTF\n\nIn-person contest\
 n24-hour Online Friday to Sat 10:00 - 18:00\n\nThe AutoDriving CTF contest
  focuses on the emerging security challenges in autonomous driving systems
 . Various levels of self-driving functionalities\, such as AI-powered perc
 eption\, sensor fusion and route planning\, are entering the product portf
 olio of automobile companies. From the security perspective\, these AI-pow
 ered components not only contain common security problems such as memory s
 afety bugs\, but also introduce new threats such as physical adversarial a
 ttacks and sensor manipulations. Two popular examples of physical adversar
 ial attacks are camouflage stickers that interfere with vehicle detection 
 systems\, and road graffitis that disturb lane keeping systems. The AI-pow
 ered navigation and control relies on the fusion of multiple sensor inputs
 \, and many of the sensor inputs can be manipulated by malicious attackers
 . These manipulations combined with logical bugs in autonomous driving sys
 tems pose severe threats to road safety.\n\nWe design autonomous driving C
 TF (AutoDriving CTF) contests around the security challenges specific to t
 hese self-driving functions and components.\n\nThe goals of the AutoDrivin
 g CTF are the followings:\n\n- Demonstrate security implications of autono
 mous driving system design decisions through hands-on challenges\, increas
 e the awareness of potential risks in security professionals\, and encoura
 ge them to propose defense solutions and tools to detect such risks.\n- Pr
 ovide CTF challenges that allow players to learn attack and defense practi
 ces related to autonomous driving in a well-controlled\, repeatable\, and 
 visible environment.\n- Build a set of vulnerable autonomous driving compo
 nents that can be used for security research and defense evaluation.\n\nTh
 e contest is based on a Jeopardy style of CTF game with a set of independe
 nt challenges. A typical contest challenge includes a backend that runs au
 tonomous driving components in simulated or real environments\, and a fron
 tend that interacts with the players. This year's contest will follow the 
 style of last year and includes the following types of challenges:\n\n- 
 “attack”: such as constructing adversarial patches and spoofing fake s
 ensor inputs\,\n- “forensics”: such as investigating a security incide
 nt related to autonomous driving\,\n- “detection”: such as detecting s
 poofed sensor inputs and fake obstacles\,\n- “crashme on road!”: such 
 as creating dangerous traffic scenarios to expose logical errors in autono
 mous driving systems.\n- “smart planner”: such as creating intelligent
  path planners for dangerous tasks that are difficult for human drivers\n\
 nMost of these challenges will be developed using game-engine based autono
 mous driving simulators\, such as CARLA and SVL. The following link contai
 ns some challenge videos\, summaries from AutoDriving CTF at DEF CON 29 an
 d DEF CON 30\nhttps://drive.google.com/drive/folders/1JSVarIaQBmseLC9Xqkfr
 xnRQto4WM225?usp=sharing (filedata/fetch?id=249421&d=1718451165)\n[URL]htt
 ps://www.youtube.com/channel/UCPPsKbVpxwk-464KIzr8xKw\n\n# What's new in 2
 024\nThis year\, we will unlock new traffic conflict scenarios that are ob
 served from real-world driving logs such as Jaywalk and double parked vehi
 cles. New difficulty levels will be added to challenges in such scenarios 
 by integrating real downstream AI modules such as object tracking from ope
 n-source autonomous driving software like Apollo\, Autoware and OpenPilot.
 \n\nIn order to enable the audience to experience the challenges more dire
 ctly\, we plan to set up a vehicle wheel controller on site and provide a 
 driving game this year. Audiences can drive themselves to compete with the
  self-driving vehicle in some of the challenges. Driving game demo:\nhttps
 ://drive.google.com/drive/folders/1LIzJJ1I3Eqj_e0_ntX5eFu82U9ObiEYB?usp=sh
 aring\n\n# For players\n- What do players need to do to participate AutoDr
 iving CTF?\nMost of the challenges do not require domain knowledge of auto
 nomous driving software or adversarial machine learning\, although knowled
 ge of those helps. For example\, the players can generate images the way t
 hey like (e.g.\, drawing\, photoshopping) to fool the AI-components or wri
 te a short python script to control the vehicle. Some challenges\, such as
  incident forensics likely would require players to learn domain knowledge
  such as sensor information format and how fusion works.\n\n- What do we e
 xpect players to learn through the CTF event?\nPlayers can (1) gain a deep
  understanding of real-world autonomous driving systems' design\, implemen
 tation\, and their corresponding security properties and characteristics\;
  and (2) learn the attack and defense practices related to autonomous driv
 ing in a well-controlled\, repeatable\, visible\, and engaging environment
 .\n\n\nTwitter: @autodrivingctf (https://twitter.com/autodrivingctf)
URL:https://forum.defcon.org/node/249895
DTSTART:20240809T180000Z
DTEND:20240811T020001Z
LOCATION:Online
END:VEVENT
END:VCALENDAR