Launched! SMART-NUS driverless car trials at One-North

- Further advances in R&D will enable the SMART-NUS driverless car to deal better with real traffic environment which includes road intersections and traffic lights, amongst others.

1.    Singapore – As one of the first applicants approved by the Singapore Land Transport Authority (LTA) for Autonomous Vehicle (AV) testing at One-North, the Singapore-MIT Alliance for Research and Technology (SMART) [新加坡-麻省理工学院科研中] and the National University of Singapore (NUS) [新加坡囯立大学], will be testing the SMART-NUS driverless car on public roads, using the 6-km route within the business park as of 12 Oct 2015. 

2.    Through this one-year trial, the SMART-NUS AV group will be advancing its vehicles’ self-driving capabilities, working towards more sophisticated demonstrations of autonomous Mobility-on-Demand (MoD) transportation services. One-North presents new challenges of road interactions with real traffic (uncontrolled environment), and more complex interactions such as cross-junctions and traffic lights. This is a progressive step for expansion of the AV research application to the entirety of the Singapore road network.

3.    Adapted from its prototype driverless golf cart, this driverless car dubbed SCOT (Shared Computer Operated Transport) has been operationally-ready for the public roads since January 2014.   Unlike other driverless cars which are retrofitted with expensive 3-D laser sensors, SCOT - a retrofitted Mitsubishi iMiEV - relies on low-cost off-the-shelf LIDAR sensors which enable the car to drive autonomously, independent of the Global Positioning System (GPS). This unique feature allows it to drive even in tunnels and places where GPS signals would be hindered.

4.    A collaborative project between SMART and NUS, this driverless car aims to resolve the “first- and last-mile problem”, which is especially pertinent in view of Singapore’s ageing society. It also aims to help promote car-sharing as the driverless car is able to resolve the ‘rebalancing’ issue, that is, getting the car to the next car-sharing customer once the previous customer drops off the car (please refer to the Factsheet).  

5.    Professor Daniela Rus, SMART Principal Investigator for the Future Urban Mobility (FM) Interdisciplinary Research Group (IRG), said: “Since the launch of SCOT in January 2014, we have added new image processing algorithms for traffic light detection; and more advanced decision-making capability for navigation through intersections such as T-junction/cross junction and zebra crossings. Our demonstration today takes us one step closer to making driverless cars a reality in Singapore. ”

6.    Associate Professor Marcelo Ang from the Department of Mechanical Engineering and Acting Director of the Advanced Robotics Centre, NUS Faculty of Engineering, added, “Going forward, our R&D will focus on how SCOT can recognise road signs, lane markings and more; and we hope that with the support of the government, we can expand this Mobility-on-Demand service demonstration over a larger section of One-North, beyond this 6-km route. This will not only help us to learn and improve the AV system, but also provide a visible platform to increase public awareness and government support in our endeavour to create better transport solutions for urban cities.”

7.    The research was funded by the National Research Foundation (NRF) Singapore under its Campus for Research Excellence and Technological Enterprise (CREATE) programme.


FACTSHEET: SMART-NUS AV RESEARCH

The objective of this research project is to allow a fleet of Autonomous Vehicles (AVs) to provide Mobility-on-Demand (MoD) which will complement the existing transportation system, so as to reduce the overall commuting time by solving the “first-and-last-mile” problem. This solution will thus reduce the traveling time from the starting location (for example, commuter’s house) to the start of the transportation network (for example, MRT station) and reduce the traveling time from the end of the transportation network to the final destination (for example, commuter’s workplace).

The research integrates existing technologies with fresh methodologies to allow driverless vehicles to intelligently provide MoD, with the goal of making this future transportation paradigm a reality. A self-driving electric car is used to demonstrate the MoD system.

The AV is a collaborative project between the Singapore-MIT Alliance for Research and Technology (SMART) and the National University of Singapore (NUS). The self-driving car has been operational since January 2014, while similar challenges have been tested on self-driving golf buggies on NUS campus since 2011.

As one of the first LTA-approved applicants for AV testing at One-North, the SMART-NUS AV group will further advance its vehicles’ self-driving capabilities, working towards more sophisticated demonstrations of autonomous MoD transportation services. One-North presents new challenges of road interactions with real traffic (uncontrolled environment), and more complex interactions. This is a progressive step for expansion of the AV research application to the entirety of the Singapore road network.

Besides solving the “first-and-last-mile” problem, AVs offer the following benefits:

Safety - Traffic accidents are the leading cause of death in road users of the 20-30 age group. Most accidents occur due to human errors. Automated vehicles have the following advantages:
1.    Quicker reaction time
2.    Shorter braking time
3.    Wider field of view
4.    Not distracted
5.    Not affected by fatigue
6.    Non aggressive behaviour
7.    Bad behaviours can be easily corrected

Accessibility - Provide mobility to people who cannot, should not, or prefer not to drive (for instance, the elderly, youth and disabled).

Productivity - “Commodity” driving is a chore that absorbs a large fraction of people’s time, which could be better used.

Efficiency/Throughput - Automated vehicles can cooperate to minimise the effects of congestion. Routes can also be planned to minimise energy wastage, such as unnecessary braking or acceleration.

Environment - Automated driving can reduce emissions by 20 to 50 per cent, and efficiently interfaces with smart power grids.

Better User Experience
1.    Routes can be planned to make the ride smoother
2.    Users can be doing other things and not worry about road conditions
3.    Possible location aware services

Electric Car Specifications (SCOT - Shared Computer-Operated Transport)
1.    3.8 (Length) x 1.6 (Width) x 1.750 (Height)
2.    Weight: ~1,200kg
3.    Max speed: 130km/h
4.    Current max. autonomous speed on roads: 30 km/h
5.    Vehicle localisation using laser sensors, and not dependent on GPS
6.    Vehicle works well in poor lighting as well as indoors
7.    Obstacle detection using laser sensors
8.    Dynamic safety zone
9.    Online booking system
10.    Total mileage on a full charge: 100-130 km
11.    Time for full charge: 6-8 hours

Enhanced Specifications (Developments in Progress)
1.    New image processing algorithms for traffic light detection
2.    More advanced decision-making capability for navigation through intersections (such as T-junction, zebra-crossings and lane merging)
3.    Road-sign and lane-marking detection
4.    Expansion of MoD service demonstration over larger section of One-North (between Fusionopolis, Biopolis, and Mediapolis).

For more, please see:
•    FM Autonomy YouTube channel - www.youtube.com/user/FMAUTONOMY

•    FM Autonomy Facebook - www.facebook.com/fmautonomy