Loading…
Transcript

TIMELINE

OneAI Versions

18.05.2023

Q3 2023

Q4 2024

Q2 2024

2025

OneAI Platform, Janine Bubner

OneAI 0.1 MVP Q2 2023

(rulebased)

Platform functionality

MVP

0.1

Backend:

continues model training in the cloud

Onboard:

OneAI native App: Service extention Container (OneAI Process inside the HU Domain together with Mbition) talks to 3 Containers (launch 3 Containers):

Basic Functions:

2 -3 Base Container (intelligence incl. IB Container & Learning Service Container (initial implementation of the IB, current running UAP on ZL kernel running within the intelligence Container), App Container (ZL tiles etc., own module control and interact with UI) add. Data & Content Container dependend on the MVP definition (collect Data from the vehicle Data interfaces)

preinstalled learning service container (dBus client/Server)

Use CASE Dummy has some interfaces to the UI and SPEECH for each individual container

between those Containers basic communication working (e.g Data can be send and App Container can use the service provider to use the Intelligence Container)

MVA Timeline? for Integration

ML Modules

Destination Prediction

Departure Prediction

ML

Cognitive Load

Team

Product App

Sina

Prasanjit

OneAI 1.0 Q4 2023

Platform functionality

OneAI

1.0

Backend

Basic Functionality MVP

Update configuration

send Data to the Backend, training model Data

Onboard

logging end to end

more Data sources like camera data for more use cases

OneAI 1.0 closer to smart intelligence e.g. simple customer modelling

Provide Interfaces to enable intelligence broker services to applications which are not running on the current OneAI platform e.g. abacus

To make rules of intelligence broker configurable

abacus can use the intelligence broker

OneAI 2.0 Q3 2024

Platform functionality

OneAI

2.0

Backend

include mobile devices into the platform - mobile platform (phone)

run a new ML module on the backend (distributed learning)

generalized federated learning (Backend Pipeline and Data collection in real time?)

ML validation (control stage deployment) before customer rollout

Onboard

abacus applications are running on the OneAI Platform

integrate all AI applications (native integration)

"Brain" Intelligence by using cognitive load, scene understanding

OneAI 3.0 Q1 2025

OneAI

3.0

platform allows newest industry cutting edge AI technology products

enabling different usecases (Platform adaption)

L4 driving usecases

edrive usecases

suspension

energy efficiency

New office

ML

Product

Expansion

FUTURE

FUTURE