Over 6.4 million automotive accidents occur in the US annually. Odds of someone being in an accident this year are 1 chance in 16. Any information that warns of problems along the road(s) ahead can therefore potentially save lives and reduce the frequency and/or intensity of accidents. The vehicle of tomorrow is the programmable-networked vehicle. In our view, the networked vehicle of the future is one of the most complex Cyber Physical Systems (CPS) with active trajectory control, active navigation and on-line maintenance. V2V wireless networks are a special class of networked-CPS where the maximum relative speeds are in excess of 80m/s, the node density can span over 9,000 vehicles/mi^2 and, most importantly, the dynamics of the vehicle, the environment, driver reaction and interaction with other vehicles need to be considered in every communication and control decision. To meet these timeliness and coordinated communication requirements, we are developing a new set of networking capabilities that can lay the foundation for dynamic vehicular networks designed to make driving safer, more efficient and more enjoyable. This project is aimed at the design, analysis, implementation and evaluation of vehicular networks that will enable a wide range of applications including V2V and V2I communication for: (a) Bounded-latency broadcast protocols for active networked safety alerts (b) Protocols and algorithms for Real-Time collision avoidance (c) Native protocols for secure V2V and V2I communication. Real-time research in V2V networks will be the first step toward developing a Spatio-Temporal Real-Time theory and network protocols with wide-area time synchronization.
The project focused on the Network Cyber-Physical Systems in the domain of automotive and transportation systems. The specific efforts were on the interaction of the car and cloud computing for (a) in-vehicle systems for remote diagnostics of automotive electronic controller units and, (b) real-time vehicle traffic congestion management. This project involved three Ph.D. students, 4 undergraduates and produced two MS thesis. The project included leadership from two female undergraduate students. Findings: 1. AutoPlug: Open Automotive Architecture for Plug-n-Play Services In 2010, over 20.3 million vehicles were recalled. Software issues related to automotive controls such as cruise control, anti-lock braking system, traction control and stability control, account for an increasingly large percentage of the overall vehicles recalled. We have developed AutoPlug, an automotive Electronic Controller Unit (ECU) test-bed to diagnose, test, update and verify controls software. AutoPlug consists of multiple ECUs interconnected by a CAN bus, a vehicle driving simulator which behaves as the plant model and a vehicle controls monitor in Matlab. As the ECUs drive the simulated vehicle, the physics-based simulation provides feedback to the controllers in terms of acceleration, yaw, friction and vehicle stability. This closed-loop platform is then used to evaluate multiple vehicle control software modules such as traction, stability and adaptive cruise control. With this test-bed we developed ECU software diagnosis and testing to evaluate the effect on the stability and performance of the vehicle. Code updates can be executed via a smart phone so drivers may remotely "patch" their vehicle. This closed-loop automotive control test-bed allows the automotive research community to explore the capabilities and challenges of safe and secure remote code updates for vehicle recalls management. We have demonstrated this to the Department of Transportation, Intel, Toyota, Ford, BOSCH and GM. Project Website: http://autoplug.org 2. AutoMatrix: A large-scale real-time traffic congestion modeling and vehicle routing platform Our goal is to investigate the construction, instrumentation and scheduling of time-bounded and anytime algorithms on multi-core architectures such as graphics processing units (GPUs). Our investigation focuses on time-bounded anytime algorithms on GPUs for real-time vehicle traffic congestion prediction and route assignment. To explore this, we have designed AutoMatrix, a traffic congestion simulation platform on the Nvidia CUDA-enabled GPU. AutoMatrix is capable of simulating over 16 million vehicles on any US street map and executing traffic estimation, prediction and route assignment algorithms with high-throughput. This research has the potential to extend real-time scheduling on massively parallel GPU architectures to attack a variety of data-driven and interactive algorithms with timely operation. 3. ProtoDrive: An Experimental Platform for Electric Vehicle Energy Scheduling and Control Protodrive is an experimental platform enabling rapid prototyping and simulation of electric vehicle powertrains. The powertrain is modeled at the small-scale in hardware, making it low-cost and compact enough to fit on a desk. It consists of a physical model of an electric vehicle powertrain coupled to an active dynamometer, making it possible to run the powertrain through its full speed and torque range. Protodrive will further electric vehicle development by: • Enabling rapid prototyping of novel powertrain architectures • Simulating federal drive cycles to determine a vehicle’s fuel consumption • Predicting range, with elevation data and a driver control strategy Presently the benefits of a battery/supercapacitor powertrain have been investigated. Protodrive runs a scaled version of an actual commute drive cycle with various battery/super capacitor charging/discharging schedules to maximize the battery’s life time and the vehicle’s range. Demonstration: http://protodrive.blogspot.com/2012/04/demo-day.html Select Awards: AutoPlug - Grand Prize in the World Embedded Software Contest, Korea in Nov. 2010. ProtoDrive - Third Prize in the World Embedded Software Contest, Korea in Nov. 2012. The PI was awarded the Intel Early-Faculty Career Honor in November 2012 for his efforts on the AutoPlug project. The PI was selected to speak at the NAE Frontiers of Engineering in September 2012 on "The Car and The Cloud: Automotive Architectures for 2020". Select Publications: R. Mangharam and A. A. Saba, "Anytime Algorithms for GPU Architectures", IEEE Real-Time Systems Symposium, 2011. U. Drolia, Z. Wang, Y. Pant and R. Mangharam. "AutoPlug: An Automotive Test-bed for Electronic Controller Unit Testing and Verification". Intelligent Transportation Systems, 2011. A. A. Saba and R. Mangharam, "Anytime Algorithms for GPU Architectures", Analytic Virtual Integration of Cyber-Physical Systems Workshop (IEEE RTSS) 2010. S. Diaz, H. Jain, Y. Pant, W. Price and R. Mangharam. "ProtoDrive: An Experimental Platform for Electric Vehicle Energy Scheduling and Control" 33rd IEEE Real-Time Systems Symposium (RTSS@Work). Puerto Rico, Dec 2012 R. Mangharam, "AutoPlug: An Open Experimental Platform for Automotive ECU Testing, Updates and Verification". NSF/USCAR Automotive CPS Workshop. 2011. M. Behl and R. Mangharam, "Pacer Cars: Real-Time Traffic Shockwave Suppression" IEEE Real-Time Systems Symposium (RTSS), WiP. 2010. A. A. Saba, S. Mohan and R. Mangharam. "Anytime Algorithms for Multicore Architectures" in 22nd Euromicro Conference on Real-Time Systems, Work-in-Progress Session, (IEEE ECRTS). 2010.
