Unleashing Real-time Insights with ClickHouse_ Navigating the Landscape in 20...
From Taxisat to inLane
1. From TAXISAT to inLane: Towards Low Cost GNSS and
Computer Vision Fusion for Accurate Lane Level
Navigation and Enhanced Automatic Map Generation
2. The importance of ITS
• The Global Market for ITS Technologies is estimated
to grow to €50BN by 2020.
• Automotive Industry is one of the most innovative
sectors
– Driver Assistance
– Autonomous vehicles
3. • Sensor Fusion is essential : no sole positioning sensor covers all
requirements and constraints
• Combination of computer vision, 3D Maps and GNSS technologies are
fostering new solutions not only for driving assistance but for unmanned
vehicles
Technology trends
4. Our basic idea
Take advantage of EGNOS/EDAS and GALILEO capabilities,
augmented by hybridization with low-cost sensors (cameras,
gyroscope, accelerometer…), to improve localization
performances while reducing production costs.
+
5. Our first approach: TAXISAT
• Objective: develop a robotic driverless GNSS and SLAM
based taxi application capable to operate cost effectively,
safely and with a high reliability within private circumscribed
sites whatever their topographic configurations are.
Start date: 01/04/2012
Duration: 24 months
8. Start from generated background:
• Fusion of Visual Odometry and EGNSS signal
• EGNOS usage studies
Towards:
• Lane Level Navigation in public roads
• Local Dynamic Map generation
• High Precision Cartography generation
• Crowdsourced map updates
Solution for niche market
Solution for mass market
9. Coordinator: Vicomtech – IK4
Duration: 1 January 2016 – 30 June 2018
• Industrial partners representing the main sectors:
– Cartography & Navigation Solution: TOMTOM
– Electronics & Components: INTEL
– Automotive & Robotic Industry: HONDA
• SME specialised in GNSS: TeleConsult Austria
• Computer Vision Research
– Visual Odometry: TUE
– Object recognition: Vicomtech
– LDM, Camera to Map alignment: HONDA
• Standarization: IFSSTAR
• End Users/Stakeholders: RACC / ERTICO
• Test sites:
– Barcelona
– DITCM
10. Local Dynamic Map generation
Estimation of vehicle position Scene understanding
GNSS Receiver + IMU
EDAS
Visual Odometry
Lane detection
Traffic Sign Recognition
11. Local Dynamic Map generation
Estimation of vehicle position Scene understanding
GNSS Receiver + IMU
EDAS
Visual Odometry
Lane detection
Traffic Sign Recognition
Cloud
12. • In-vehicle computer: navigation system / driver
assistance / Autonomous Driving.
• Aftermarket device: navigation system / driver
assistance.
• Smartphone: navigation system / driver
assistance.
Target products
14. Currently…
Working on first prototype:
• Defined requirements and architecture.
• Integrating first version of SW modules.
• Preparing testing:
• Defined methodology.
• Setting up instrumented vehicle
(HW & SW).
15. Conclusions
• ITS Market represents a huge opportunity for the GNSS
sector
– From driver assistance solutions to autonomous vehicles
• Sensor cost will be a limiting factor for the technology
penetration
• Sensor fusion is the key for the next generation
affordable high automated vehicles
– Precise and enriched Cartography is the first step towards
the autonomous vehicles
• Standards: need to develop GNSS performance
standards to ease the deployment of ITS application
based on GNSS solutions
Close circuits – dedicated lanes
Not embedded solution
Not mass market target
Target Markets: Airports, campus, logistics etc.
Limited investment maximum reliability
Niche market: autonomous vehicle for close circuits with dedicated lanes, to use it in airports, parks, and so on.
OEMs and other players in the automotive industry are looking for lower-cost methods to construct and maintain maps.
So we decided to move from a solution for a niche market to a solution for mass market that could answer to the necessities of the automotive industry: to create lower-cost methods to construct and maintain maps.
Niche market: autonomous vehicle for close circuits with dedicated lanes, to use it in airports, parks, and so on.