Rising Demand for Vehicle Data and Its Impact on Mobile Computing, Security and Privacy Research
Kang G. Shin
Kevin & Nancy O’Connor Professor of Computer Science, Department of Electrical Engineering and Computer Science, University of Michigan, Ann Arbor, MI, USA
Date & Time
09:00 ~ 10:15 January 18th (Monday), 2021.
Connected vehicles are estimated to generate 25 GB of various data per hour per vehicle, including data from cameras, LiDARs, radars, etc. The in-vehicle network (IVN) itself, consisting of multiple network buses, such as the Controller Area Network (CAN), produces only a fraction of this amount of data (~200 MB/hour in newer vehicle models), but still carries highly valuable and critical sensor information. The generation and sharing of this driving data are anticipated to create an additional source of revenue for OEMs and third-party services of up to $155.9B by 2022. OEM-independent, universal access to data by third-party service providers can make the latter a major player in automotive data monetization.
Vehicular data is not only a growing target for monetization by the above commercial solutions, but also enabling academic and basic research in mobile computing, security and privacy. For each of these three research areas, this talk will cover some of our ongoing work: DETROIT, LibreCAN and Spy.
To accommodate the growing need of vehicular data collection and processing, we have been developing an open-source vehicle-agnostic end-to-end framework for vehicular data collection, translation and sharing (DETROIT) to facilitate the rapid development of automotive apps. Besides the support of various vehicle interfacing tools through a Vehicle Hardware Abstraction Layer (VHAL) and data accessibility by third-party service providers, it can also translate raw un-interpretable CAN data whose semantics are proprietary and a secret to the OEM.
The translation is done by a framework called LibreCAN which can automatically translate most CAN messages with minimal effort (using 30 minutes of free driving data and 10 minutes of data collected in a closed experimental setting). The tool is designed to save academic researchers the time and effort they would otherwise spend manually reverse-engineering the CAN messaging format of each vehicle they study. Our data collection tool can detect the make, model and year of the current vehicle, check if a reverse-engineered translation table exists in its backend database and kick off the process of collecting the required CAN data to run LibreCAN on. During regular operation of the data collection tool, i.e., sharing data with a third-party offline or online/real-time, the data will then be translated on the fly using the translation parameters obtained through LibreCAN.
Besides the usefulness for academic vehicular data collection, LibreCAN also has benefits towards automotive security. Vehicle security attacks to date have all shared one very important feature – they all ultimately require write access to the CAN bus. But in order to do that, one has to know the message format of the CAN bus to inject meaningful data. All makes and models of vehicles have different message formats that are proprietary to the car manufacturer which wishes to prevent cybersecurity attacks on vehicles by not disclosing translation tables for CAN data. In order to cause targeted and intentional changes in vehicle behavior, malicious CAN injection attacks require knowledge of these translation tables. As a result, security researchers can determine how new attacks can be used against a number of makes and models at once and design necessary defenses quicker.
Besides the facilitation of mobile computing and security research, as well as data monetization in the automotive industry, vehicular data collection will also lead to the study of privacy of the collected data due to an increasing awareness of emerging privacy concerns. Under the current suggested permission model of commercial automotive telematic/infotainment platforms, we find those leaking users’ location information without explicitly obtaining users’ permission. In a novel privacy attack called SPy, we found a high accuracy of inferring a vehicle’s location from seemingly benign steering wheel angle (SWA) traces, and showed its impact on the driver’s location privacy.
This is joint work with Mert Pesé, a PhD student of Real-Time Computing Laboratory of University of Michigan.
KANG G. SHIN is the Kevin & Nancy O'Connor Professor of Computer Science in the Department of Electrical Engineering and Computer Science, The University of Michigan, Ann Arbor. His current research focuses on QoS-sensitive computing and networking as well as on embedded real-time and cyber-physical systems, such as autonomous vehicles.
He has supervised the completion of 86 PhDs, and authored/coauthored close to 1,000 technical articles, a textbook and about 60 patents or invention disclosures, and received numerous awards, including 2019 Caspar Bowden Award for Outstanding Research in Privacy Enhancing Technologies, and the Best Paper Awards from the 2011 ACM International Conference on Mobile Computing and Networking (MobiCom’11), the 2011 IEEE International Conference on Autonomic Computing, the 2010 and 2000 USENIX Annual Technical Conferences, as well as the 2003 IEEE Communications Society William R. Bennett Prize Paper Award and the 1987 Outstanding IEEE Transactions of Automatic Control Paper Award. He has also received several institutional awards, including the Research Excellence Award in 1989, Outstanding Achievement Award in 1999, Distinguished Faculty Achievement Award in 2001, and Stephen Attwood Award in 2004 from The University of Michigan (the highest honor bestowed to Michigan Engineering faculty); a Distinguished Alumni Award of the College of Engineering, Seoul National University in 2002; 2003 IEEE RTC Technical Achievement Award; and 2006 Ho-Am Prize in Engineering (the highest honor bestowed to Korean-origin engineers).
He has chaired Michigan Computer Science and Engineering Division for 3 years starting 1991, and also several major conferences, including 2009 ACM MobiCom, 2008 IEEE SECON, 2005 ACM/USENIX MobiSys, 2000 IEEE RTAS, and 1987 IEEE RTSS. He is the fellow of both IEEE and ACM. He has also served or is serving on numerous government committees, such as the US NSF Cyber-Physical Systems Executive Committee and the Korean Government R&D Strategy Advisory Committee. He has also helped founding a couple of startups.
Practicing the Art of Data Science
Professor, School of Computing Science, Simon Fraser University, Burnaby, BC, Canada
Date & Time
11:15~12:30, January 19th (Tuesday), 2021.
Data science embraces interdisciplinary methodologies and tools, such as those in statistics, artificial intelligence/machine learning, data management, algorithms, computation and economics. Practicing data science to empower innovative applications, however, remains an art due to many factors beyond technology, such as sophistication of application scenarios, business demands, and the central role of human being in the loop. In this talk, I would like to share with the audience some experience and lessons I learned from my journey in data science research and development. First, I will illustrate the core value of building domain-oriented, end-to-end data science solutions that can help people gain new interpretable domain knowledge. Second, using network embedding as an example, I will demonstrate that the nature of data science practice is to connect challenges in vertical applications with general scientific principles and tools. Last, I will advocate data supply chain, which connects data supply and consumption based on value of data and value-added services.
Dr. Jian Pei is working hard to facilitate efficient, fair, and sustainable usage of data and data analytics for social, economical and ecological good. Through inventing, implementing and deploying a series of data mining principles and methods, he produced remarkable values to academia and industry. His algorithms have been adopted by industry, open source toolkits and textbooks. His publications have been cited more than 100,000 times. He is also an active and productive volunteer for professional community services, such as chairing ACM SIGKDD, running many premier academic conferences in his areas, and being editor-in-chief or associate editor for the flagship journals in his fields. He is recognized as a fellow of the Royal Society of Canada (i.e., the national academy of Canada), the Canadian Academy of Engineering, ACM and IEEE. He received a series of prestigious awards, such as the ACM SIGKDD Innovation Award, the ACM SIGKDD Service Award, and the IEEE ICDM Research Award. Currently he is a full professor at Simon Fraser University, Canada.
Autonomous systems in smart cities
Professor, Dr. rer. nat., Department of Computer Science, Hamburg University of Applied Sciences, Germany
Date & Time
17:00 ~ 18:15 January 19th (Tuesday), 2021.
Almost every larger city in Europe has ambitious smart city projects. This is particularly true for Hamburg, a Hanseatic city in the north of Germany. Hamburg is the smartest city in Germany according to a Federal Association for Information Technology. Although there are no megacities in the European Union (the largest city in the European Union is Berlin with 3.7 million inhabitants), the increasing urbanization is apparent and produces problems to be solved. At the same time rural depopulation creates conjugated problems.
One category of these problems is mobility. Mobility can be regarded as the need to move persons and freight. In densely populated cities an increasing amount of transport users have to share a decreasing amount of space with conflicting needs. At the same time in rural areas, a dwindling supply of local public transport makes the mobility of the remaining residents more difficult. The same applies to parcel delivery or the supply of goods.
Autonomous systems have great potential to create a sustainable and livable environment. The author has initiated a publicly funded project to investigate technologies of autonomous mobile systems which interact with a smart city. The test area intelligent urban mobility (Testfeld intelligente Quartiersmobilität) at the campus of Hamburgs University of Applied Sciences is created to do research on connected and autonomous mobile systems like multipurpose robots and other mobility users like pedestrians with a smartphone. A particular focus is on neighborhood mobility. This means that distances of less than 3 kilometers usually have to be covered.
The special type of needs in neighborhood mobility has two important aspects that affect development of autonomous mobile systems: It is slow mobility and the transport users are especially vulnerable. The acceptance of the residents of autonomous systems is equally important, as is the protection of privacy when collecting environmental data.
That being said, there are technological difficulties in developing real autonomous systems.
They are expected to make decisions on their own in complex environments. The real world usually differs from a simulation or an experimental setup in a laboratory – a problem commonly referred to as Sim-2-Real gap. Active and non-destructive exploration is expected from an autonomous system to solve unexpected problems. Machine learning methods come into play which in turn have their own pitfalls.
The author has built a specialized laboratory to investigate machine learning technology applied to autonomous systems. In this laboratory miniature autonomous vehicles are developed. The general idea of this exerimental setup allows research on new methodologies for autonomous systems in a very small scale.
Stephan Pareigis has studied applied mathematics at Munichs Ludwig-Maximlians-University. He has written his Ph.D. thesis on numerical methods for optimal control problems and reinforcement learning. Stephan has more than a decade experience in developing software for distributed and embedded real-time control systems in various
companies. Since 2004 Stephan holds a professorship of Applied Mathematics and Technical Computer Science at Hamburg University of Applied Sciences. His research interests are in autonomous systems, real-time and reactive programming, embedded machine learning and reinforcement learning. In particular Stephan has been developing small robotic systems and autonomous miniature cars for more than 15 years. He has been supervisor for student teams participating in autonomous systems competitions. Together with his autosys research group, Stephan is currently setting up a publically funded test field for smart mobility in a city district in Hamburg. Stephan is currently dean of the Department of Computer Science at Hamburg University of Applied Sciences.
ⓒ Copyright 2021 KIISE – All Rights Reserved.
[KIISE] Korean Institute of Information Scientists and Engineers
#401 Meorijae Bldg., 76, Bangbae-ro, Seocho-gu, Seoul 06704, Korea
Email: email@example.com | Fax: +82-2-521-1352 | Homepage: www.kiise.or.kr
Business Registration Number: 114-82-03170