Singapore And Autonomous Vehicles — Interesting Lessons In Governance, Planning, And Safety
Singapore is a dense city-state with a large growing population (approx 6M). In 2015, the Singapore government launched the Sustainable Singapore Blueprint(SSB) which had five strategic thrusts to build a sustainable city, and one of the key elements was “a car-lite” Singapore. Given the limited space, it is no longer possible to increase road infrastructure or cars without significantly adding to congestion. Thus, SSB focuses on shifting mobility to walking, biking, and public transportation.
Why look at autonomous vehicle technology?
Given the increasingly point-to-point nature of travel for both passenger and logistics, there is a tradeoff between large buses (perhaps on inefficient fixed routes) and smaller shuttles with a more dynamic responsive transportation system. This need for point-to-point is amplified for a rapidly ageing population.
“The challenge is that both public transport as well as logistics (internet shopping and therefore related home deliveries are increasing) are highly constrained by the availability of drivers,” said Niels de Boer, Programme Director, Centre of Excellence for Testing & Research of Autonomous vehicles at Nanyang Technological University, Singapore (NTU Singapore).
In 2018, Deputy Secretary of Land and Corporate at the Ministry of Transport, Lee Chuan Teck, published a far-reaching and visionary article “Public Transport Planning and the Technological Revolution.” Critical elements of this article were:
- Freeing public transportation from fixed routes and schedules
- Using Autonomous Vehicles(AV) to overcome manpower constraints
- Narrowing the gap between private and public transportation
Most critically, there was a recognition that new technologies create new challenges. “AVs, for example, require financial support and regulatory sandboxes to facilitate experimentation,” said Deputy Minister Teck. Also, the report recognizes the potential shift in labour. “ For example, when autonomous buses are ready, bus drivers will no longer be driving buses. They may instead be trained to handle exceptions, such as when the autonomous buses break down, or commuter incidents, “ says Deputy Minister Teck.
How does one turn vision into a plan which ensures the safety and builds a framework for regulation?
Similar to Jacksonville, a joint team was developed between CETRAN, the Land Transport Authority, traffic police and JTC(planning agency). CETRAN is the Centre of Excellence for Testing & Research of AVs operated by engineers from NTU — one of the world’s top-rated universities. Thus, a critical linkage was forged between operators, regulators, and technology experts. The Land Transport Authority falls under the Ministry of Transport, and the overall initiative falls under the Smart Nation Program which is run directly from the Prime Minister’s office. Given the broader potential impacts of AVs beyond transportation, the CARTS (Committee for Autonomous Road Transport Singapore) was created to include representatives from other ministries, agencies, and also included industrial representatives.
Quite sensibly, the objective of the framework was to support trials with a testing regime which links an increase in the test area based on demonstrated capabilities. Any vehicle currently running on public roads has gone through the CETRAN AV Test Centre to assess its readiness in terms of safety for such trials. Currently, all testing requires an onboard safety driver. This structure gives assurance to the public relative to safety and provides a structure for the regulator to learn about the technology and build effective long term regulation. As Ike report points out, Artificial Intelligence (AI) machines can have robustness issues and an inclination to “learn only the test.” Thus, the details of the framework tests are not publicly available. In 2019, the area for the testing framework was expanded to Western Singapore.
In terms of funding, the Singapore government has awarded funding for AV trials to a number of organisations such as APTIV (Delphi, NuTonomy), Singapore Technologies and NTU-Volvo. In January, they opened a CFC (Call for Collaboration) which includes phased development of an R&D phase, trial phase and pilot program phases.
In terms of progress to date, the core group has worked with 10 different developers of autonomous vehicles and more than 30 vehicles have been tested for the purpose of getting an AV authorisation which allows them to be trialled on public roads. The largest fleet is from APTIV, but there are also developers who have just a single vehicle running. Vehicles tested are anything from small cars based on Renault Zoe and Mitsubishi iMiEV to full-size 12m city buses from NTU-Volvo and ST Engineering Land Systems.
The results have offered insights on AV testing and evaluation. An interesting one comes from the traffic police. “In the driver’s licence exam, the traffic officer looks for indicators and behaviours which indicate that a driver has observed risk and is taking action to mitigate that risk,” said Mr. de Boer. How does one make such an assessment of an AV? Is it important? As mentioned in SAE EDGE Research Reports, there is an increasing recognition that transportation systems have a “language of driving.” A critical part of this language is the communication of risks and intent.
Overall, public transportation in the major cities of the world may have the biggest need for AV technology. One suspects that these cities will be the front-lines for technology development, regulatory learning, and eventually proliferation.