00:07:38 Joash Lee: Good morning from Singapore =) 00:17:09 Ziran Wang: @Joash Welcome! 00:33:28 Ziran Wang: @Everyone Please feel free to add your questions to speakers here in the chat box. Thanks. 00:34:30 Jia Hu: （@zhanbo) what's the difference when you mentioned "closed loop" and "hybrid"? 00:37:25 XB Hu: @zhanbo how far into the future do you need to predict? 00:37:58 Nigel Williams: @Zhanbo will the slides be posted later? 00:39:23 Xishun(Heeson) Liao: @Zhanbo for trajectory prediction, do you consider surrounding vehicle's behavior into your model? what kind of information are needed? 01:01:22 Nigel Williams: @Haoran how do you decide which gap the rear CAV should cut into? 01:05:38 Sergei Avedisov: @Haoran, how many vehicles realistically can platoon and make simultaneous lane changes like this? Have you looked at platoon length vs mobility results? 01:21:30 Ziran Wang: @Everyone Please feel free to ask your questions to the lightning talk speakers in this chat box as well. Thanks! 01:24:44 Xinpeng Wang: @Xishun For highway merge, how is right-of-way represented in the utility function design? Thank you! 01:25:14 Nigel Williams: @Heeson do manual vehicles still play the game? 01:25:25 Sergei Avedisov: @Xishun, have you looked at penetrations that are small, like 5% or 10% 01:26:20 Xinpeng Wang: Got it. Thank you! 01:27:40 Sergei Avedisov: Followup, when do you start seeing a significant difference? 01:29:15 Xishun(Heeson) Liao: You are welcome!@Xinpeng 01:35:56 Sergei Avedisov: @Yougang, what gain ranges would you expect to tune your CAV to? 01:36:27 Sergei Avedisov: In terms of alpha/beta 01:38:23 Sergei Avedisov: Would some of those values be unreasonable for vehicle dynamics constraints? 02:24:23 sistla: how does the vehicle detection in the side lane happens? 02:28:58 sistla: So is this similar to sensor fusion of Radar + Lidar + v2x messages ? 02:35:04 sistla: Do you have any pointers on measuring channel resources 02:36:57 Xinpeng Wang: @Sergei Really interesting talk! In NGSIM data, error goes to zero using CPM at 30% penetration rate. Is there a math guarantee on how much penetration needed to achieve no error for all such scenario? (It sounds similar to herd immunity.) 02:38:28 sistla: This is really interesting talk , can you share the papers and reference link to your paper 02:38:47 Jia Hu: Yeah... couldn't find the paper either 02:40:33 Xinpeng Wang: Good to know. Thanks! 02:44:25 sistla: Ziran can you provide me link for your webpage to access these 02:44:41 Ziran Wang: Sure. Here it is: https://ziranw.github.io/iv2021workshop/ 02:45:17 sistla: Thank you Ziran. This is really interesting 02:45:30 Ziran Wang: Thank you for attending!