Differences

This shows you the differences between two versions of the page.

--- gsoc:2024-gsoc-agl [2024/01/12 18:31]
till created
+++ gsoc:2024-gsoc-agl [2024/03/13 13:41] (current)
jsmoeller
@@ Line 15: / Line 15: @@
 Wiki – http://wiki.automotivelinux.org
-Documentation (including getting started guide): https://docs.automotivelinux.org/en/needlefish/#
+Documentation (including getting started guide): https://docs.automotivelinux.org/en/master/#
 Git - gerrit.automotivelinux.org
@@ Line 26: / Line 26: @@
 Primary Mentor contact: jsmoeller (at) linuxfoundation (dot) org
+GSoC 2024 intro presentation: https://docs.google.com/presentation/d/1eo4Veo5b7HkNgRnL1MWfZcXNj89Hh5RdeBDSBf6bEaQ
 ===== Project Proposals =====
-**To be updated**
-  * Beginner (short/half-time/175h):
+==== Task #1 (large/350h):   Extend AGL's demo control panel ====
-    * Task #1a: Write a flutter demo application (tbd)
-    * Task #1b: Write a html5 demo application (tbd
+The AGL demo control panel is a pyQt (qml) application that is able to
-    * Contact: jsmoeller (at) linuxfoundation.org
+control the AGL demo homescreen.
+  * add file playback capability
+  * visualize 4 signals (realtime and/or from playback)
+  * stretch-goal: add IC keypad feature
+  * stretch-goal: add power-control feature
+Level of difficulty: intermediate
+Requirements: QT, pyQT, QML, grpc, kuksa.val
+Contact: jsmoeller (at) linuxfoundation.org
+==== Task #2a (large/350h): Extend AGL's speech recognition app (Flutter app) ====
+The AGL speech recognition app is a Flutter app that uses a speech recognition engine (vosk-kaldi)
+for the audio processing.
+This task is about adding the whisper (https://openai.com/research/whisper) framework as well.
+Level of difficulty: intermediate
+Requirements: python, flutter+dart
+Contact: jsmoeller (at) linuxfoundation.org
+==== Task #2b (large/350h): Extend speech recognition app with AI chatbot / AI intent engine ====
+The AGL speech recognition app is a Flutter app that uses a speech recognition engine.
+It could be extended with an AI chatbot or an AI intent engine to execute API calls.
+Level of difficulty: intermediate
+Requirements: python, flutter+dart
+Contact: jsmoeller (at) linuxfoundation.org
+==== Task #3 (large/350h): Persistent storage API ====
+This task is about to develop an grpc-API for AGL that serves as persistent storage API for the demo using e.g. sqlite as backend and another local engine (pluggable backends).
+Level of difficulty: intermediate/difficult
+Requirements: C/C++ or rust, grcp
+Contact: jsmoeller (at) linuxfoundation.org
+==== Task #4 YOUR OWN IDEA ====
+Bring your own idea to extend the project.
-  * Intermediate:
+Level of difficulty: tbd
-    * Task #2a (long/full-time/350h):
-      * Write an application in flutter //or// Qt5 //or// HTML5 that simulates CAN bus traffic using  kuksa.val (aka like a simulator for can messages)
-      * Update documentation pages based on work.
-      * jsmoeller (at) linuxfoundation.org
-    * Task #2b (long/full-time/350h):
+Requirements: tbd
-      * Create a setup with the "Carla" Simulator to be used to write CAN frames to the AGL demo unit.
-      * Update documentation pages based on work.
-      * Contact: jsmoeller (at) linuxfoundation.org
-  * Advanced:
+Contact: jsmoeller (at) linuxfoundation.org
-    * Task #3 (long/full-time/350h): Platform Development
-      * Replicate the containerized IC-EG demo with KVM (aka from system containers to KVM)
-        * Tasks include work on Yocto, KVM passthrough topics and
-      * Contact: jsmoeller (at) linuxfoundation.org
-  * Advanced:
-    * Task #4 (long/full-time/350h): App + Platform Development
-      * Enhance the speech recognition using vosk
-      * This requires an update to the userspace libraries used in the platform
-      * And also requires updating a flutter app

Wiki

User Tools

Site Tools

Differences

Page Tools