P16241: Autonomous People Mover Phase 3
/public/

Integrated System Build & Test with Customer Demo

Table of Contents

Team Vision for System Level Demo with Customer

The team accomplished a major milestone for this phase. The team received in the main circuit board and tested its functionality. The circuit board proved to be everything and more than what was planned. It cleaned up a lot of the noise in the responsiveness seen in the control subsystems. Because of these results, the team was able to take the cart outside and test the remote control mode. This testing displayed extraordinary results and can be seen in the videos on the page.

The team also continued working on obstacle detection which included looking into the LiDAR and ultrasonics in order to be able to process the data into usable information for obstacle detection.

Furthermore, the team mounted all the needed components on the enclosure panel along with re-wiring the cart to make sure no loose hanging wires were present and to ensure proper connections.

The team also gave a presentation at the 2016 St. Lawrence ASEE Conference at Cornell on April 9, 2016. This presentation included giving an overview of the autonomous people mover as well as the engineering process and effort that went into making the cart what it is today. This presentation can be seen here.

Test Results/Project Progress

Full Controls Test

The team tested the control systems while the cart was still on blocks. This test involved testing the functionality of the steering, brake, and throttle systems. Furthermore, these systems were applied in intervals of 50 and 100% showing that these systems can be applied in intervals rather than full application. The video of this test can be seen below.

Outside Testing

The team took the cart outside to test the functionality of the new circuit board. The team tested this using the remote control to test the responsiveness of the control subsystems like steering, braking, throttle, and the emergency stops. The video of this outside testing can be seen below.

Main PCB

The team received in the PCB and quickly got to soldering the components on to check its functionality. When tested, the PCB showed extraordinary results, displaying exactly what was expected. The PCB was then mounted into the enclosure along with many other components. Pictures of the PCB can be seen below.
Finished PCB mounted on panel with all necessary components in place

Finished PCB mounted on panel with all necessary components in place

Finish PCB mounted on panel in enclosure

Finish PCB mounted on panel in enclosure

GPS

The GPS underwent multiple tests trying to maintain a fix on several satellites in order to receive a kinematic fix. This deemed to be very difficult and varied greatly, however, when there was a fix on several satellites the results were astonishing. A picture can be seen below showing the accuracy of the GPS when establishing a constant fix from several satellites.
GPS test showing communication with satellites along with the accuracy of the GPS

GPS test showing communication with satellites along with the accuracy of the GPS

LiDAR

For the LiDAR, Hector_Slam was used in order to build a map while pushing the LiDAR around the MSD floor area. This could be used for building maps for the cart to follow when out in the outside environment. This would be key in autonomous driving in the fact that it would continuously update maps to account for things that have moved or changed from the trip that was taken at a given area. Pictures of this map can be seen below.
Hector_Slam map built from recorded LiDAR data of the MSD area with a red line showing the direction the LiDAR was pushed

Hector_Slam map built from recorded LiDAR data of the MSD area with a red line showing the direction the LiDAR was pushed

Another Hector_Slam map built from recorded LiDAR data of the MSD area with a red line showing the direction the LiDAR was pushed

Another Hector_Slam map built from recorded LiDAR data of the MSD area with a red line showing the direction the LiDAR was pushed

IMU

The team started looking into the IMU to be able to obtain odometry data for the cart in order to help with localization. The team did this by enabling DMP on the MPU-9250 and verified that the results received were correct. The team will then implement these results into ROSdue in order for the cart to read in the values to help localize itself.

Risk and Problem Tracking

Risk Management

The team is still monitoring and adding risks as they arise while also closing out risks that are no longer in effect. This document can be seen below.

Problem Tracker

The team is still recording all the issues that arise and what the action plan is to fix them or deal with them. This document can be seen below.

Budget Information

The team put together a document that entailed all of the purchases for Phases III, II, and I. This document can be seen below.

Functional Demo Materials

Plans for next phase

For the next phase, the team plans on focusing first on localization and allowing the cart to realize its relative location and which direction it is headed. Secondly, the team will focus on developing a map of the course area for the cart to follow using input points and possibly Hector_Slam. Next, we will start plotting points on the map for the cart to know which areas are valid and invalid. Lastly, the team will focus on getting the ultrasonics and LiDAR to function as obstacle detectors to ensure that while on the path, the cart will not hit any objects.

Home | Planning & Execution | Imagine RIT

Problem Definition | Systems Design | Subsystem Design | Preliminary Detailed Design | Detailed Design

Build & Test Prep | Subsystem Build & Test | Integrated System Build & Test | Integrated System Build & Test with Customer Demo | Customer Handoff & Final Project Documentation