王飞跃的个人博客分享 http://blog.sciencenet.cn/u/王飞跃


主编引语:平行驾驶+交通5.0 + 未来交通

已有 5652 次阅读 2014-12-30 18:06 |系统分类:观点评述

Finally, before the end of the first day of 2015, I am able to finish all remining 10 papers for assignment and 2 papers for final decision, all left from 2014. Techically, I still keep my promise to compelete all papers within the year since it is still 2014 in USA.

In 2014, I was late for 3 papers by three days, and 3 papers by one day, all for assignment, 1 paper for one week for decision. This is not bad since I have processed about 800 new submissions this year, plus over 500 papers left from 2013 and 2012.

IEEE Transactions on Intelligent TransportationSystems, Vol.15,No.4, pp1381-1387

Scanning the Issue and Beyond: Parallel Driving WithSoftware Vehicular Robots for Safety and Smartness

Parallel Driving With Software Vehicular Robotsfor Safety and Smartness

Recently, autonomous driving and intelligentvehicles have become hot topics again. However, the reality is that the automobileindustry is one of the last few sectors that has not yet been revolutionized bythe Internet technology. We do see a lot of guarded gates, roads, and bridgesfrom big automotive corporations and their allies, such as OnStar,G-BOOK, SYNC+AppLink, iDrive, inkaNet, as well as CarPlay, Android Auto, Windows in theCar, etc., but this is far from the vision of a world of connected vehiclesand infrastructures, as least far from the vision outlined in many papers ofour special issue and our white paper “ElectrifiedVehicles and the Smart Grid: The ITS Perspective,” the first articlepublished in this issue.

In 2003, I wrote a technical report entitled“Parallel Vehicular Services for Intelligent Vehicles: From Real Cars to ArtificialCars” to promote the idea of a distributed framework for control and managementof advanced driver assistance systems (ADAS) based on mobile agents thatconnect cars, homes, offices, and service centers, which was able to conduct variousservices for operators under the principle of “simple inside vehicles, complexoutside vehicles” for design and the implementation of connected vehicularplatforms. From there, we proposed and launched the first IEEE InternationalConference on Vehicular Safety and Electronics in the next two years with Prof.Nanning Zheng (then the President of Xi’an Jiaotong University) in order topromote further research and applications in this direction. The idea wasfunded partially by China’s Ministry of Science and Technology, in 2004, as akey project under its “863 Program,” in which we developed a few keytechnologies and a vehicular application specific operating system (vASOS) for real-time embedded softwareplatforms based on OSEK/VDX and OSGi protocols. Since then, parts of the reporthad been published in journals and conference proceedings, and a number ofproposed directions became the topics for Ph.D. dissertations of my graduatestudents, from 2003 to 2007, and led to a few patents. I had actually stopped workingin this area for a few years until three years ago. I was reinvigorated mainlyby the significant progress made in autonomous driving by Google and the newpublicity brought to the field. Originally, I had thought these leaps would be impossiblewithin the next 50 years due to potential legal hurdles created by lawyers inthe United States. With the recent advances in mobile communication, knowledgeautomation, robotics, and artificial intelligence techniques (particularlymachine learning and natural language processing) as well as the increase inpublic awareness and excitement on smart cars over Google’s autonomousvehicles, I believe we must reevaluate and adjust our own approaches anddirections for research and development in intelligent vehicles. To me, now isthe time for various smart technologies to begin incorporating into cars and autoservice centers. I would like to see technologies in the form of softwarerobots driving and working in parallel with human operators, so that operatorscan more effectively and easily keep their attention on roads and cars andenjoy a safe, smooth, and smart driving experience.

My original idea was very simple: developingtask-specific agents for various driving and maintenance/service functions thatcan be hosted and executed by simple local but real-time platforms inside avehicle. These agents could also move to complex remote but powerfulenvironments outside the vehicle, where they can be enhanced in performancethrough further training and learning with human or computer programs. After readingChristopher Steiner’s Automate This: How Algorithms Came to RuleOur World two years ago, I think now is the perfect time to expand thisidea and make those agents smart web surrogates or software robots for bothautonomous driving and ADAS. To this end, we need to rethink and redefine many ofour current vehicular hardware and software systems, such as ABS, ACC,DSC/TCS/ESP, lane departure warning, pedestrian detection, obstacle avoidance,fatigue detection and warning, smart speed adaptation, etc., viewing them asvehicular robots, making them mobile through networks and able to learn from driverbehaviors and local environments by cloud computing. In this way, these systemscan be adaptive and refined over time and individually optimized for the bestperformance. Of course, we need a more open platform and many more specified protocolsin order to develop more such software robots to perform diversified tasks:monitor the state of drivers, check the condition of cars, evaluate the safetyof environments, and identify the availability of services during drivingprocesses. We also need to establish reputable test beds that can thoroughly evaluate,verify, and certify these vehicular robots, so that people can create, test,and buy them easily with confidence. Of course, security and privacy couldstill be a big issue for such scenarios of parallel driving.

I believe this is not a dream too far fromreality. In fact, this year I have graduated my first Ph.D. in parallel drivingresearch. Although the technical orientation and depth of the dissertation wasmore basic, this is a good start toward parallel driving with softwarevehicular robots and achieving safety and smartness in the age of driving incyberspace.

FEI-YUE WANG, Editor-in-Chief


IEEE Transactions on Intelligent TransportationSystems, Vol.15,No.5, pp1861-1868

Scanning the Issue and Beyond: Computational Transportationand Transportation 5.0 

Computational Transportation and Transportation5.0 

For the past few days, I have spent most of myfree time reading and thinking about a book entitled INDUSTRIE 4.0, puttogether by a few German experts; the book is a follow-up to Recommendationsfor Implementing the Strategic InitiativeINDUSTRIE 4.0, the finalreport of the Industrie 4.0Working Group, published earlier last year as aneffort to secure the future of German manufacturing. I received the book inJuly as a gift from SIEMENS China after my lecture on ACP-Based Smart ParallelManagement of Intelligent Enterprises at the SIEMENS Industrial Forum forFuture Manufacturing. To be frank, I was quite disappointed with the book’scontent—it did not cover or explore much new ground academically—and was verysurprised by the lavish praise it has received so far from the media andso-called experts. I am really not sure whether those experts had actually readthe publication or if they were knowledgeable about the field. Many of theideas and suggestions outlined in the original INDUSTRIE 4.0 have been publishedin AdvancedManufacturing Technology in China: A Roadmap to 2050, for which I wasone of the coeditors. Written in 2007 and made public in 2009 as one of 18volumes in Science & Technology in China: A Roadmap to 2050, I hadthe opportunity to be a part of the project as a leading organizer andcontributor. In my reading, the key point of the entire book can be distilledin a passage found in the beginning of Chapter 7 by Anton S. Huber of SIEMENS:“. . .this can be understood as inside the inner part of an enterprise, there existsa parallel digital enterprise.” This is an idea I have long championed—if Ihave more time, I would like to write a review article titled INDUSTRIE4.0—Queen’s New Clothes.

Reading the book brought back some old memoriesfrom my early research career and inspired some new thoughts on the future oftransportation research. I started my official research career in fracturemechanics (yes, I did have an “unofficial” research career before that) andthen moved into computational mechanics. I left the field of computationalmechanics after one particular experience: it had taken me an entire month to locatea single bug in my Fortran program. In the middle of thousands of lines, therewas one highly unfortunate typo—the letter “l” replaced with the with number“1” in a variable name! Note that only very primitive line debugging tools wereavailable in the early 1980s, instead of graphic-based interactive integrateddebugging environments today. As educational as the month-long experience was,I was left depressed at thoughts of what I could have done with thatmonth—better used for reading and writing instead of looking for such aridiculous and meaningless bug among lines and lines of codes. However, myfailure attempt at computational mechanics left me with a strange but strongcomplexity with fields that started with the term “computational.” This mightbe what led me to coin the terms “computational sociology” and “computationalsocial studies,” what led to research in my social computing and computationalsocieties over a decade ago. I am glad that those fields have emerged as majornew directions around the world today. Now, I think it is time to promote thefield of Computational Transportation as a research direction to integrateand lift the current work in computer simulation and computational analysis oftransportation systems to a new and higher ground, where powerful new computingmethods, advanced sensing techniques, and big data in cyber, physical, andsocial spaces can be easily utilized—much like what has happened incomputational mechanics, computational fluid dynamics, computational physics,computational chemistry, computational social studies, and many othercomputational X. I am glad to see a few conferences and educational programs onComputational Transportation Science (CTS) have emerged over the past fewyears, and the University of Illinois at Chicago has even started a Ph.D.program in CTS.

From Computational Transportation, we can leap toTransportation5.0! Why? I will not discuss the definition of Transportation 1.0 to3.0, but Cyber-Physical Systems (CPS) has been claimed to be the foundation of INDUSTRIE4.0. Transportation is a direct result of human and social activities; thus,we must have social component explicitly represented in transportation researchand development. Therefore, my vision of the next generation of intelligenttransportation systems must be based on Cyber-Physical-Social Systems (CPSS), astep beyond CPS and, thus, INDUSTRIE 4.0, consequently, Transportation5.0. Yes, currently, we are still in the early stage of CPS-basedTransportation 4.0, as witnessed from connected vehicles to electrifiedinfrastructure.

As a veteran in real-time embedded systems, I wasexcited to see the emergence of CPS a decade ago as the next generation ofnetworked embedded systems, with focus shifted from hardware to software. In2008, as the Associate Editor-in- Chief of IEEE Intelligent Systems, Iproposed to launch a new department called CPS Department and promised to writethe first article for its inaugural issue. However, I felt the coming ofreal-time social signals, mainly from my own blogging experience, thus the needto add social dimension to embedded applications. After becoming the magazine’sEditor-in-Chief in 2009, I decided to change the name of the department from CPSto CPSS and started the first issue in 2010 with my article “The Emergence ofIntelligent Enterprises: From CPS to CPSS,” which overlapped INDUSTRIE4.0 significantly in key ideas, technical processes, and systemarchitectures, but with explicit and fundamental considerations for connected andreal-time social and human aspects. Now, with Facebook, Twitter, Weibo, and, inparticular, Wechet, real-time social signals, particularly real-time socialsignals for transportation (socalled social transportation), arepervasive and easily usable. I have seen some emerging revolutions in China:taxi drivers

in metropolitan Beijing have become connectedhuman (not a program or an algorithm) optimizer, constantly thinking,calculating, and optimizing their routes, actions, and profits based on real-timesocial signals received from smartphones and onboard cab-sharing informationsystems, social signal-based public transportation systems in Chongqing andother cities, and, of course, the significant behavior changes in ordinarycitizens when they are calling a cab or taking a bus (a senior citizen, who canbe less savvy with smartphones, finds that it is difficult to hail a taxi onthe street now, a problem we need to address soon). There are even moredramatic changes to come if all cars are as connected and roads as electrifiedas mobile phones and distributed energy traders and storage devices are now, asoutlined in our whitepaper “Electrified Vehicles and the Smart Grid: The ITSPerspective” (see vol. 15, no. 4, pp. 1388–1404).

It is my belief that CPSS-based Transportation 5.0,more specifically, software-defined transportation systems, O2O (online tooffline and vice verse) computational transportation experiments, and paralleltransportation with knowledge automation for closed-loop control and managementwith society-wide feedback, will be realized for real-world application beforewe have the ability to know and see what an actual CPS-based Transportation 4.0system is. 

FEI-YUE WANG, Editor-In-Chief 


IEEETransactions on Intelligent Transportation Systems, Vol.15,No.6, pp2153-2159

Scanning the Issue and Beyond: The T-ITS Awardsand Future Transportation  

The T-ITS Awards and Future Transportation

First, presenting the Best Editor Service Awards(formally called Best Associate Editor Award). As usual, three editors havebeen selected, but the final decision was an extremely tough one. Since 2009,when I became the Editor-in-Chief, the number of manuscripts submitted annuallyto our Transactions has increased from 322 to 675, our page count has increasedfrom 720 to 2772 pages, and the number of papers published per year increasedfrom 68 to 234. However, our average evaluation timelines has greatlydecreased! From submission to the first decision and the final decision, ourtimes have been reduced from 156 days to 71 days and from 273 days to 103 days,respectively. The average number of manuscripts processed annually by eacheditor is now 14, a record high that must be significantly reduced as soon aspossible. This is a tremendous demand on time and effort from our editors! I reallyappreciate their service and dedication, and this is why it has been such adifficult decision to make for this year’s Best Editor Service Awards.Congratulations to the following three distinguished editors:

QingjieKong, Chinese Academy of Sciences and Qingdao Academy of Intelligent Industries(QAII);

Rosaldo Rossetti, University of Porto;

Mauro Da Lio, University of Trento.

Again, many thanks for their outstanding efforts andgreat contributions to our Transactions.

Here, I would like to call for more volunteers forassociate editors and reviewers. Anyone with two papers published in ourTransactions is qualified to be an associate editor. Please send your CV alongwith your two T-ITS published papers to me at feiyue.trans@gmail.com.

Second, the Best Paper Awards of the IEEE TRANSACTIONS ONINTELLIGENT TRANSPORTATION SYSTEMS. The Award Selection Committee has selectedthe following two papers for the Best Regular Paper Award and the Best SurveyPaper Award among those published by our Transactions during 2012 and 2013.

Best Regular Paper Award

A. Hofleitner, R. Herring, P. Abbeel, and A. Bayen“Learning the Dynamics of Arterial Traffic From Probe Data Using a DynamicBayesian Network,” IEEE Trans. Intell. Transport.Syst., vol. 13, no. 4,pp. 1679–1693, 2012.

Best Survey PaperAward

     Li Li,Ding Wen, Nan-Ning Zheng, Lin-Cheng Shen “Cognitive Cars: A New Frontier forADAS Research,” IEEE Trans. Intell. Transport.Syst., vol. 13, no. 1, pp. 395–407, 2012. 

Congratulations to the Authorsof both papers!  

Finally, I would like to say a few words on the future ofthe Transportation. The ultimate goal of our T-ITS publication is to promoteand advance the development and deployment of effective and efficienttransportation technology. This year marks the 100th birthday of the trafficlight application. According to History.com (http://www.history.com), on August5, 1914, the first electric traffic light was installed on a city street inCleveland, Oh, USA, marking a major milestone in the history of trafficmanagement. Ten years later, in 1924, Siemens installed the famous five-sidedtraffic light tower in Berlin’s Potsdamer Platz, the busiest intersection inEurope at that time. For better or worse, the red–yellow–green signals are nowan integral part of urban landscapes worldwide. What will be the future oftraffic management in another 100 years? With or without traffic lights? I hopenot!

I do like the idea of Siemen’s historical “Traffic Light Tower,”not as a singular traffic regulator for one intersection, but re-imagined forthe future as a fully connected transportation computing and decision-makingbuilding or buildings supported with the future generation of the Internet ofEverything and cloud computing for real-time, closed-loop, and adaptive urbantraffic management. Each city would have its own “Transportation Tower.” Inthis future, we may no longer need traffic lights, with a fleet of smartvehicles, linked and autonomous, able to share pervasive, multimodetransportation information available from physical, cyber, and social domains viaroadside devices, mobile equipment, and human sensors. My vision of theTransportation Tower will be a typical CPSS, that is, a cyber-physical-socialsystem, a full implementation of our vision of Transportation 5.0, as discussedin my previous editorial, published in October 2014. For this, I am so gladthat our call for a Transportation 5.0 Initiative has received an overwhelmingresponse in a very short time. At the 2014 ITSC in Qingdao, we launched a jointinternational research center for intelligent transportation (hosted andsupported by Qingdao Academy of Intelligent Industries) to facilitate thedevelopment of our vision and recommendations for a CPSS-based Transportation5.0 Strategic Initiative. I will present more information on this excitingeffort in my editorial next yea

I am looking forward to driving into the future ofintelligent transportation. To all of our readers, a very Happy New Year! 

FEI-YUE WANG, Editor-in-Chief


上一篇:[转载]《自动化学报》40卷12期网刊已经发布, 敬请关注, 谢谢:)

2 李伟钢 李本先

该博文允许注册用户评论 请点击登录 评论 (1 个评论)


Archiver|手机版|科学网 ( 京ICP备14006957 )

GMT+8, 2018-5-28 09:32

Powered by ScienceNet.cn

Copyright © 2007-2017 中国科学报社