The overall goal of the project is targeted to improve the Singapore Milestone Framework evaluation of Autonomous Vehicles. The focus areas were pursued with the aim of improving the milestone framework and have a follow-on impact on improving the overall capability of autonomous vehicles for deployment in Singapore’s urban tropical environment. They are on the areas:

• Increasing the ability to evaluate perception systems and improving requirements on simulation-based assessment. Perception systems could be evaluated based on placement of sensors, and further understanding of sensor performances within Singapore’s dynamic urban tropical environment. The integration of perception stack within simulation explored and utilized a open-source simulation framework as reference simulation platform for ADS testing. This allows further refinement in the assessment of AV behaviour to incorporate realistic traffic interactions such as random traffic, and perception capability within simulation environment.
• Increasing the ability to objectively assess AV behaviours on roads which includes interaction with vulnerable road users. These vehicle behaviours apply across simulation, circuit, and on-road assessment as well as post deployment monitoring. These capabilities new methods are developed through a driving risk evaluation tool, defining complexity of on-road interactions to challenge AVs, and understanding of human perception of AVs at road crossings which provides insight for the refinement and recommendations of expected AV behaviours.

Multiple collaborations with partners within the AV industry were made during this project such as AV developers, software providers on simulation and accident and incident database, international bodies such as ITU FGAI4AD.

Connected Smart Mobility (COSMO) programme is a national R&D initiative hosted by Nanyang Technological University (NTU) in collaboration with ASTAR’s Institute of Infocomm Research (I2R), Centre of Excellence for Testing & Research of AVs (CETRAN) and Singapore University of Technology and Design (SUTD) to develop next generation connected mobility technologies and solutions.
Supported by National Research Foundation’s IAF-PP funding programme, we strive to build a holistic connected mobility industry ecosystem by aligning and integrating public sector research with early industry involvement and collaboration.
COSMO programme comprises 4 R&D tracks with 8 work packages that provide a comprehensive suite of technologies, solutions, and best-of-breed V2X mobility use cases that enhance road safety, optimise traffic and support future smart mobility solutions globally as well as to realise Singapore’s Smart Mobility Vision 2030.

Autonomous Driving Simulation & Validation is an international project, carried out in close coordination with the French Institute of Research and Technology System X and its industrial partners Renault Group, Systra, SNCF and AV simulation.
The overall objective of the project is to develop and demonstrate a simulation-based testing framework for the evaluation and validation of the performance, reliability and safety of autonomous vehicles to support authorizations for public operation.

• Compare and analyse existing methods and tools, provide guidelines and illustrate a validation framework on simulation for autonomous vehicle systems, taking into account all stakeholders.
• Create a Proof of concept (POC) for a simulation platform generating scenarios and their variations with the aim to develop a method to assess the safety performance of the ADS based on the measurement of a set of metrics.
• POC using open sources technologies for ADS (Baidu Apollo, Autoware.AI ) and for AV simulator (Carla and SVL simulator).
  

The end goal of SIEMENS V2X set up is to deploy a digital driver for Autonomous bus shuttles assisted with an on-road perception system. This infrastructure based perception will provide a precise picture of the traffic situation, which would enable autonomous vehicle to make better decisions without the need for human personnel on board, but still ensuring Safety & Efficiency.

In addition to aiding autonomous vehicles, such systems will provide quality information which could be used for road traffic control and management, thereby reducing the accident rates and congestions.

This system is being designed and tested with the help of a digital twin which allows it to be validated in a virtual environment before actual trials are deployed for testing and full service.

Siemens is in collaboration with ERI@N to explore the system design and validation techniques for infrastructure assisted Autonomous vehicles to realize applications including traffic light status communication and control along with extended on-road perception for Autonomous Vehicles.