THE PROFILE 2017/18

Mobility/Autonomous Vehicles/3D Vision

Photo: Bombardier

In cooperation with industrial partners such as Bombardier, AIT is developing unique driver assistance systems.

AUTONOMOUS

VEHICLES

The development of intelligent driver assistance systems and technologies for autonomous vehicles, aircraft and transport and work machinery using 3D technologies is one of the most important industry-related AIT research areas.

OUR SERVICES


3D reconstruction of environments

Object recognition, classification and tracking

Self-localization

Path and motion planning


PROJECTS


INDUSTRIAL PARTNERSHIPS

Driver Assistance System for Trams, Client: Bombardier

3D Dental Scanner, Client: a.tron3d

Embedded Stereo Vision on FPGA, Client: IEE S.A.

Augmented Reality and Indoor Navigation

Human Detection and Obstacle Avoidance for Mobile Machines


AUTONOMOUS LAND VEHICLES

SafeCon – Safe Semi-Autonomous Convoying *

RelCon – Reliable Control of Semi-Autonomous Platforms *

SWAMO – Sensing and Reconstructing the World Around Mobile Machines (grant-aided as part of FIT-IT)

AnyView3D – combining multiple visual cues for robust 3D sensing (grant-aided as part of BRIDGE)


AIRBORNE VISION

SkyObserver – unmanned aircraft systems for airborne reconnaissance in emergency and disaster management *

HWS-PAS – Highway in the Sky Pilot Assistance System **

VASL – Visual Airborne Self Localization **

DVKUP – Collision avoidance systems for pilot assistance **

* grant-aided in the KIRAS Safety Research Program of the Federal Ministry for Transport, Innovation and Technology (BMVIT)

** grant-aided as part of TAKEOFF

OUR APPROACH


SELF-LEARNING IT SYSTEMS

AIT has extensive expertise in the field of innovative computer vision and self-learning IT systems in the research area of „Intelligent vision systems“. In this context, AIT Experts have developed a special 3D sen- sor technology that allows reliable and driver-independent 3D capturing of the environment in real-time by vehicles and is therefore an essential component of vehicle assistance systems. The development objectives are increasing safety by assisting the driver, and improving efficiency, especially in the field of mobile work machines in industry, construction and agriculture. These technologies form an important basis for fully autonomously operating vehicles.


RAIL, ROAD AND OFF-ROAD

The AIT technologies are developed in close collaboration with industry partners to produce production-ready systems. A case in point is the co- operation between AIT and Bombardier on developing a driver assistance system for trams, a global first. The tram learns to recognize various ob- jects, e.g. vehicles and persons, in advance to independently assess their danger potential for possible collisions and to react accordingly, thus rev- olutionizing tramway safety. In the field of construction, the innovative as- sistance technology is moving into the field of heavy mobile work ma- chines in cooperation with Liebherr, capturing the vehicle surroundings to reduce the risk of accidents for drivers and workers.


FULLY AUTONOMOUS VEHICLES

The AIT technologies also enable the realization of fully autonomously operating vehicles. In 2016, a modern drive-by-wire tractor was presented as part of the „European Land Robot Trial (EL-ROB)“; it was later devel- oped further into a completely autonomous vehicle for use in special of- froad scenarios using special AIT driver assistance technology – an impor- tant step towards the development of new, promising fields of application e.g. in agriculture (precision farming). This technology development is car- ried out in close cooperation with Austrian partners from industry as well as the BMLVS (Austrian Federal Ministry for National Defence and Sport) and is funded by the Federal Ministry of Transport, Innovation and Technol- ogy (bmvit) within the research program „ICT of the future“.


AIRBORNE VISION

In the aviation sector as well, sensor technologies developed at the AIT enable the detection of possible obstacles, both in the air and on the ground (e.g. objects on the runway). This enables realising assistance systems for pilots or autonomous flying systems. This requires a relia- ble obstacle detection under different environmental conditions, the ability to independently perform evasive maneuvers in order to avoid collisions and a secure self-localization of the aircraft, even indepen- dent of satellite navigation. A key objective of the developments is to in- crease safety in the manned aviation by supporting the pilot as well as to expand the application possibilities of unmanned systems, e.g. for the transport of goods in hazard zones or compiling the situation pictu- re in crisis management and disaster management, e.g. in case of fires, floods, avalanches or major events, or to locate vulnerable persons or determine the spread of hazardous gases. The monitoring of critical in- frastructures such as power lines, dams or industrial plants also repre- sent important future deployment scenarios.

Photos: AIT/Michael Mürling, AIT/Zinner, Lunghammer – TU Graz

The self-driving tractor was one of the highlights at the ELROB 2016.

Prof. Dr. Thomas Pock, Principal Scientist am AIT Center for Vision, Automation & Control

DI Dr. Manfred Gruber, Head of Business Unit Autonomous Systems