Linux Embedded Challenge Projects - 2014

Yes

1. Adaptive Dynamic Headlights with Pedestrian Spotlight Function - by eVision

Team members:
- Balaban Valeriu - Master, Advanced Microelectronics, Electronics, UPB
- Voicu Tudor Alexandru - Bachelor, Applied Electronics, Electronics, UPB
- Stanescu Sebastian - Bachelor, Telecom Networking and Software, UPB

Short description:
Because of the high rate of fatalities between night accidents a lot of research are held for developing techniques to
increase the driver area of vision during the night and to reduce the accident damage if this could not be avoided. The
adaptive headlight function helps to see further in poor light conditions and especially in bends: the cornering light
swivels the headlights in the direction of travel, with the degree of turn computed by a CPU, to illuminate as much road
area as possible
An interesting solution is spotlight lighting function, which is a LED beam that specifically illuminates potential hazards.
If the near infrared camera detects deer at the roadside or pedestrians on the road, they can be briefly illuminated
beyond the normal area covered by the main beams, by a spot-light to attention the driver for a possible danger.

Code Sources
- https://github.com/izzi/app-evision
- https://github.com/izzi/meta-evision

2. Voice Commanded Interface for DriverVehicle Interaction - by She#

Team members:
- Iulia Neagoe - Computer Sciences and Military Information Systems,Military Technical Academy
- Mihaela-Anca Sorostinean - Computer Sciences and Military Information Systems,Military Technical Academy

Short description:
In the context of continuous technological advances in the automotive and communications domains, a drivers
responsibility has switched from just controlling the car to interacting with the multitude of gadgets provided by the
manufacturer. The purpose of this project is to design an interface that offers the driver the possibility of controlling
some of the nonvital functionalities of the car by vocal commands, so as to enable the driver to focus his attention
on the road while also having a comfortable mean of communication with the car.
We developed a system of vocal recognition of some basic features like the radio, windows, clima or a phone
which we implemented on the Wandboard. Also we provide the user with a graphical interface of the recognized
commands in order to enhance his interaction with the vehicle.

3. Driving Control Software - by FreeSoftwares

Team members:
- Petrosanu Adrian-Sabin - Computer Science, UPB
- Birsan Nicoleta Cosmina - Computer Science, UPB
- Radoi Ioana Gabriela - Computer Science, UPB

Short description:
"Driving Control Software" is a soft to control an automatic gearbox. This project consists of simulating the behavior
of an automatic gearbox on the Wandboard. An automatic gearbox is a type of motor vehicle transmission that can
automatically change gear ratios as the vehicle moves.

4. Autonomous Car Parking - by ATM

Team members:
- Mihai Coca - Computer Sciences and Military Information Systems,Military Technical Academy
- Georgian Andrei - Computer Sciences and Military Information Systems,Military Technical Academy
- Hiji Iulian - Computer Sciences and Military Information Systems,Military Technical Academy

Short description:
A large number of companies are developing autonomous vehicle technology through applying its work in the area to
a particular usage case : parking. The purpose of this project is to design a concept vehicle, which can be dropped off
at the curb by its owner and left to its own devices to enter into a spot park. The process can even be reversed
when the owner is ready to go, with the car leaving the spot park on its own to meet its key-holder again at the curb.
Documentation:
- Please consult ATMDoc.pdf

5. Collision Detection - by Beer2.0

Short description:
The purpose of our project is to provide cars with a sense of the road ahead and enable it to take preventive actions against
collisions; In this way we hope to reduce accidents on the road. It will gather signals and informations from different hardware
and will alert the driver or take immediate control of the car in order to make critical maneuvers in order to protect the driver
from any life threatening event.
A Freescale Cup car will be equipped with a Wandboard and two USB cameras so we can track the environment. After initial
object tracking we will incorporate Human Interaction by remote control. For this project we believe a simple warning system
and/or breaking will suffice as a proof of concept.

6. Feedforward adaptive noise cancellation using sub-band normalized filtered-X LMS algorithm - by Brainiacs

Team members:
- Cristian Monea - Telecommunications and Information Technology, Electronics, UPB
- Madalin Zaharia - Telecommunications and Information Technology, Electronics, UPB

Short description
This project proposes a feedforward adaptive noise cancellation (ANC) algorithm based on sub-band normalized filtered-X LMS (NFXLMS).
The use of an adaptive algorithm offers advantages over simple filtering algorithms like fixed FIR or IIR filters. Also, noise generated in a
car environment can be considered stationary because it preserves some of its properties, like spectral distribution, mean, variance, which
allows the use of adaptive filters in car noise cancellation applications.
The feedforward system should be more efficient than feedback systems. In this case, a coherent reference noise input is sensed before it
propagates past the canceling speaker.
Thus, the algorithm will simulate the two sensors (microphones): reference sensor, which measures the primary noise to be canceled,
and the error sensor.