ECE 492 Project Research Proposal By: Stephen Beggs

Need For New Project As electric vehicle approaches completion, we need to look into the next big project for the department New project should retain positive qualities that the LFEV project had Enough work to occupy a whole class of ECE’s Both Digital and Analog Solutions Be relevant to industry trends What can we change? Make the project “mean” something Collaborative opportunities

Criteria Requires both digital and analog solutions Properly prepare students for post-grad professional life by immersing them in an industry-realistic environment Reflects the values of Lafayette College Gives the project a “so what” Holds relevance to current events/has potential to produce impactful results Limited search based on large projects aiming to solve large real world problems

Final Options Wind Power Facial Recognition Technology Self-Driving/Autonomous Vehicle Technology Further Research into Electric Cars GPS Technology

Self-Driving Cars Requires both Digital and Analog Solutions Relevant to current industry trends Young technology that our students could potentially contribute to Has significance to students, school, and Lafayette community as a whole Opportunity for collaboration Enough work to occupy ECE senior class for multiple years Requires team building/communication skills Project reflects values of Lafayette College

Building a Self-Driving Vehicle System A LOT of work to be done System can be broken down into several main subsystems Localization/Mapping Obstacle Avoidance Path Planning Control System team Interconnect team

Localization/Mapping Vehicle needs to know where it is on the earth, as well as where it is in relation to its surroundings Laser Rangefinders/Lidar/Radar Allows car to perceive how far it is away from the objects around it High definition Camera Allows car to perceive color and shapes System combines these perception techniques to create 3D map Team needed to put localization system Together Use GPS unit to help car perceive its position on earth

Obstacle Avoidance Once vehicle can perceive its surroundings, it must be able to react to what is around it Internal Map Static and predicted locations of objects around it Objects determined based on cross referencing incoming data with predetermined sets of shape and motion descriptors Descriptors are used to determine what is what and where is it going in the world around the vehicle Vehicle must react properly according to what it sees around it

Path Planning How to safely direct the vehicle from point A to point B Avoiding accidents Following rules of the road Use of GPS unit in cooperation with localization system to help vehicle plan optimal path Determine safest path by analyzing path options and current status What commands need to be given to the car once the path has been chosen? Series of throttle, brake, and steering commands

Control System/Interconnect In order for a vehicle to perceive its surroundings, know where it is on the earth, react to obstacles, and plan an optimal path, there must be a central control system Aggregate incoming data from sensors and GPS unit Receive input from obstacle avoidance system and react accordingly Receive input from path planning system and react accordingly Make sure all subsystems can talk to each other as efficiently as possible Interconnect team needed to physically connect all the systems together

Collaborative Opportunity Serious ethical/policy based implications Opportunity to work with policy study and/or gov/law majors Based on potential policy issues and/or new policies what kind of restraints should be put on car? Rules of the road? How safe is safe enough? What kinds of policies should we implement to protect the passengers of self-driving vehicles from being hacked? Opportunity to combine engineering minds with non-engineering minds to accomplish meaningful and impactful research Collaboration opportunity provides realistic work setting for students Collaboration reflects school values/mission

Heilmeier’s 9 Critical Questions What are you trying to do? Articulate your objectives with absolutely no jargon. What is the problem? Why is it hard? How is it done today, what are the limits of the current practice? What’s new in your approach and why do you think it will be successful? Who cares? If you’re successful what difference will it make? What impact will successes have? How will it be measured? What are the risks and payoffs? How much will it cost? How long will it take? What are the midterm and final “exams” to check for success? How will progress be measured?

Questions?