DENSO www.densocorp-na.com Collaborative Automated Driving Description of Project DENSO is one of the biggest tier one suppliers in the automotive industry, and one of its main goals is to provide solutions to the OEMs in the Autonomous Driving field. DENSO plans to increase its contribution to the development of automated driving technologies by enhancing the functionalities of existing techniques and invent new solutions. The main objective of the project is to develop new methods to enhance the capabilities of automated system by providing extra information to help the system make proper driving decisions. This technique will increase the perception range of the system - beyond the field of view and line of sight of the local sensors - by sharing the information via the wireless medium. The goal of the project along with high level description of the main components are summarized as follows: Introduction Collaborative Automated Driving (CoAD), via sharing sensory information across vehicles, is an emerging technology in Autonomous Driving (AD) field. This technology has received high attention from researchers in both industrial and Academic sectors. The main goal of CoAD is to enhance automated driving by assisting vehicles to increase their perception range to go beyond the Field of View (Fov) of their local sensors. CoAD can also provide increased redundancy to the AD system. The shared information can be used to enhance localization and make proper driving decisions. In this project we will study the techniques of object detection and sharing and investigate the impact of information sharing concept on enhancing automated driving functionalities (e.g., Localization and path planning). Information sharing will be studied considering two driving scenarios. The wireless communication technique that will be used for information sharing and message exchange is Dedicated Short Range Communications (DSRC). The main tasks that will be addressed in this project are Object(s) Detection and Representation Map Creation CoAD Message Profile A high level overview of these tasks is presented in the following sections. Object(s) Detection and Representation Autonomous vehicles are equipped with an array of sensors for environment perception. These sensors help the vehicle detect, locate, track the movements of other vehicles and extract their dynamics. However, due to wireless channel (DSRC) limitations raw or even processed data Page 1 of 7
cannot be shared over the channel. Therefore a presentable set of information should be extracted from each detected object and sent out to the air. Two cases will be considered under object detection and sharing as follows: Moving and Static Object Representation The objects that can be detected and tracked by vehicle s local sensors can be divided into two types: static (e.g., Landmark, traffic signs, traffic light locations, etc.) and moving objects (e.g., vehicles, trucks, bicycles, pedestrian, etc.). Some of the landmarks can be obtained from High Definition (HD) maps but those maps are not accurate all the time, which has bad impact on the localization functionalities. In this task, we will study and define different methods for object representation taking into account DSRC limitations. Figure 1 depicts an example of the objects that can be detected and how they can be represented for wireless sharing. FIGURE 1. TYPES OF OBJECTS THAT CAN BE DETECTED AND THEIR REPRESENTATION Map Creation One of the main challenges of autonomous driving is detecting construction zones and navigating through detours around the zone, especially when there is no updated map of the area. In this project, we will investigate the techniques of creating a local map of the detour and share it with other vehicles. The map can be created with the help of different sensors; Camera, LiDAR, internal sensors and GPS. Figure 2 shows an example of a vehicle navigating through a detour that spans multiple lanes. In this example, a manual driven vehicle or automated vehicle creates a local map of the detour as it navigates through the cones. Receiving multiple and different copies from different vehicle helps the autonomous vehicle build an accurate map (Continuous sharing and matching will result in an accurate map of Page 2 of 7
the area). When approaching a construction zone, the AV will be able to localize itself and navigate safely through the constructions (The more information about the map and its surroundings the more enhanced localization). FIGURE 2: A MAP OF A DETOUR AROUND THE CONSTRUCTION ZONE Map modeling procedures can be studied from the perspective of map creation (on the transmit side) and map re-construction and matching (on the receive side) A map can be modeled and represented using a set of functions and some key objects around the detour. Vehicles create path history points as they move between the cones, connect them in a piecewise function(s) fashion, and calculate confidence levels for these points. The functions along with the key points and the confidence levels are shared. Map fusion and matching should also be studied and proper methods for fusion and matching should be presented. CoAD Message Profile There is a tradeoff between communication performance and the quality of data; and therefore the CoAD message profile (e.g., size, data rate, transmit rate and power, etc.) should be carefully designed to meet the application requirements. When designing a message profile, different factors, like channel limitations, communication environment, delivery range and other factors, should be taken into consideration. A min set of information should be used to represent objects so that receiving vehicles are able to recognize and locate them on the map. In this task, an optimal message profile should be defined, and should be designed such that sensory information are carried and delivered in all driving conditions. Communication Protocol Vehicle-to-Vehicle (V2V) communication technology will be used for this project. The focus should be on the App layer (no need to study the PHY layer) to be able to generate wireless messages. The students should understand the Basic Safety Messages (BSM) structure and should be able to modify those messages (e.g., adding/deleting new fields, re-ordering the fields). I strongly recommend studying SAE 2735 standard (focusing on BSM section) and understand the overall concept of V2V safety communication Page 3 of 7
DENSO will provide the student team with A) SW architecture in a block diagram form. The diagram shows how App layer is linked to DENSO SW services through APIs. The implementation is done using C language under Linux OS. B) At least two physical units. Each unit is capable of sending and receiving data (mainly BSMs). A document with instructions on how to connect those units to laptops will also be provided. An alternative approach is using simulation and DENSO will provide you with the simulation platform ahead of time. Include a Deliverable (Phase I) and Details Here: BASELINE GOAL Scope Problem understanding and an analysis of how the final solution is envisioned. Defining proper methods that will be used for feature extraction, object representation, map creation, and data fusion. Define the precision for each object type Show system limitations in terms of object representation, map creation and communication capabilities. define the features that will be extracted from the objects and best use of them Define metrics and the evaluation process Include a Deliverable (Intermediate Phase II) and Details Here: SUCCESS Scope CoAD final structure Implementation. Min/max number of objects that can be shared, the order of these objects and the order of the elements allocated for each object. Considering extra information (e.g., accuracy, confidence, etc.) for each object and show how this information can help identifying the objects on the receive side. Map implementation (creation and fusion). Preliminary results Include Stretch Goals and Details here: HIGH SUCCESS Scope Show how AD functionalities (like Localization) can be improved using CoAD. Show how continuous sharing improves map accuracy and how AD benefits from the map for localization and navigation. Page 4 of 7
Location Most of project work will take place on campus during the semesters Project Sponsor Mentor Zaydoun Rawashdeh: Received his Ph.D. degree in Electrical and Computer Engineering from Wayne State University, Detroit, USA, in 2011. He is currently with DENSO Research and Development North America leading the Connected-Autonomous Driving project. His research experience includes V2V-based active safety systems, Object matching and tracking, Embedded Systems, and Mobile Ad Hoc Networks. He has published several papers in connected vehicles domain. He is a member of the Society of Automotive Engineers (USA), the Institute of Electrical and Electronics Engineers, Inc. (USA), Rajesh Malhan: Received his PhD. Degree from the University of Delhi, India in 1989. He was postdoctoral research fellow in the Department of Electrical and Electronic Engineering, Toyohashi University of Technology, Japan till Mar. 1991. He Joined DENSO in 1991. Now, he is the director of advanced research at DENSO s North American Research and Development Department. He is responsible for DENSO s Automotive (Autonomous drive, Vehicle electrification, Vehicle wireless charging, Alternative fuel and Metal air battery technologies) R&D work. He has published more than 50 research journal papers, inventor or co-inventor on over 25 US and 150 International patents, and contributed chapters in 4 books. He is a winner of 24th JSAP (Japan Society of Applied Physics) year 2002 award for outstanding achievements in the field of applied physics. He is a member of the Society of Automotive Engineers (USA), the Institute of Electrical and Electronics Engineers, Inc. (USA), the Japan Society of Applied Physics (Japan) and served on many international professional society committees. Project Faculty Mentor Prof. Wei Lu, Mechanical Engineering Research interests include Modeling and simulation of nano/microstructure evolution; mechanics in nano/micro systems; mechanical properties and performance of advanced materials and the relation to their microstructures. Key Skills & Project Roles Areas of Experience: Computer Networks, Artificial intelligence, Control, Robotics Page 5 of 7
Student Role Likely Majors (1) Artificial Intelligence/Machine Learning Electrical Engineering, Computer Engineering, Computer Science (CSE/CS-LSA), Robotics (2) Control Systems Electrical Engineering, Computer Engineering, Computer Science (CSE/CS-LSA), Robotics, ISD-AUTO (1-2) Wireless Communications Electrical Engineering, Computer Engineering, Computer Science (CSE/CS-LSA) (1-2) Data Analysis MIDAS / MICDE Certificate Program Graduate Students, STATS, Industrial & Operations Engineering Desired skills: Strong analytical modeling and programming (C, C++, Matlab, etc.), Knowledge of network simulators is also important Company Overview DENSO is a leading supplier of advanced automotive technology, systems and components for major automakers. DESNO has committed to making the world better place through its worldfirst products and technologies. Page 6 of 7
More information about DENSO can be found at the following link: http://www.globaldenso.com/en/about-us/at-a-glance/ Legal Requirements This project is open to all students regardless of citizenship status Intellectual Property Agreements / Non-Disclosure Agreements (please select) Students will sign the standard MDP IP/NDA agreement Internship Information Summer Funding Available Page 7 of 7