In 2013, Karl Iagnemma, former director of the Robotic Mobility Group at MIT, allied with Emilio Frazzoli, Director of the Transportation@MIT initiative, to create Nutonomy, aiming to tackle the most difficult challenge in self-driving cars, urban driving. In August 2016, this startup attracted wide public attention by beating Uber to become the first company to publicly test self-driving cars in the world.
Leaders League. With so much competition from traditional automakers and tech giants, why do you think a startup like nuTonomy succeeded in launching the first autonomous taxi service in the world?
Karl Iagnemma. A key differentiator is the fact that the company’s heritage is robotics, not traditional automotive production. nuTonomy is pioneering technology for motion planning and decision making in its software system that is based on methods that have been used successfully in the development of spacecraft, planes, and other complex, safety-critical autonomous systems. Formal logic, which is used in spacecraft, enables nuTonomy’s vehicles to be more reliable and robust in their features, as well as drive in a more human way.
Our combination of state-of-the-art sensors and algorithms is able to monitor and react to a vast range of situations presented by other vehicles, pedestrians, cyclists, and other road users. We have built our software specifically to navigate unpredictable scenarios that may occur on the road.
What has been the feedback of different stakeholders following your Singapore and Boston launches?
Our ongoing public trials in Singapore continue to provide us with valuable feedback from riders on the entire experience - from booking to arriving at the final destination. Reactions from riders so far have been very positive. People begin the rides hesitant or excited, but seem to quickly accept the fact that the car drives like they or any other driver would, and the anxiety dissipates. We’ve also seen passengers humanize the car by giving it a name or comparing its driving to people they know.
We’re pleased with the initial testing program that city and state officials have launched, and we look forward to continuing to work with the City of Boston to expand the testing area for nuTonomy vehicles in the future.
What are your suggestions to develop autonomous vehicles (AV) in the US?
A significant uncertainty facing nuTonomy is whether we can win approval to test our vehicles in the range of urban environments that are required in order to develop a safe, efficient, fully autonomous mobility-on-demand transportation service. In the US to date, decisions about where and how AVs may be tested have been left to individual states. The US Department of Transport’s model state policy was helpful, but federal regulators may need to take further steps to ensure that states do not impose undue burdens on AV testing.
We expect that the Trump administration will continue the Obama’s administration’s support for AV testing and development. We encourage the new administration to do more to develop the conditions for the safe testing and rollout of this rapidly advancing technology. To this end, we have three suggestions for the administration to build on the clear-eyed policy guidance found in the Federal Automated Vehicles Policy:
First, AV regulations should encourage fair competition. The potential economic impact of autonomous vehicles has attracted the interest of a diverse set of businesses, ranging from the world’s largest corporations to a few engineers in a garage. Some organizations have multi-million dollar lobbying budgets and armies of lawyers to help them comply with burdensome regulations. Others, like nuTonomy, are startups that do not.
Regulation should not create barriers to market entry. As with some federal regulation, the compliance burden should be sensitive to the scale of regulated companies. For AVs in particular, the number of miles or hours a company has driven should not be a critical metric, since such a metric favors organizations that test large AV fleets primarily on highways, where it is easy to rack up miles. Regulators should instead derive metrics focused solely on the performance of a company’s technology. Such an approach would let the best companies – large or small – win regulatory approval.
Second, regulation should not slow the pace of technical development. At nuTonomy we update our software continuously, based on data we collect during on-road testing and large-scale simulations. Our AVs can safely negotiate traffic situations today that they could not tackle only weeks ago. NHTSA’s Vehicle Performance Guidance contains a request that companies notify regulators “four months before active public road testing begins on a new automated feature.” NHTSA should streamline this notification requirement as it could have the unintended consequence of delaying the deployment of safer versions of AV technology. The cost of such unnecessary delays would be measured in lives lost behind the wheel.
Third, government should help ensure that liability risks associated with AVs are manageable. We at nuTonomy believe that if and when failures in AV technology lead to injury, businesses should be held responsible. However, unpredictable verdicts and damages awards can dry up insurance markets and ultimately make our nation’s roads more dangerous. In the 1980s, Congress helped to promote another life-saving technology – vaccines – by establishing the National Vaccine Injury Compensation Program to recompense victims fairly and efficiently. This approach keeps vaccine manufacturing economically viable, and could serve as a model for the next administration’s approach to the emerging AV industry.
Singapore seems to provide an ideal environment for innovation. What do you think other countries can learn from it?
The population of Singapore is tech-savvy and highly connected. Connectivity is key to nuTonomy's growth, since we rely on access to a digital platform to e-hail an autonomous vehicle. nuTonomy has also benefited by recruiting from the strong computer science talent base in Singapore.
nuTonomy was the first company in the world to test its autonomous vehicles in a public trial, and we were able to do that in Singapore. Singapore as a nation has been very supportive of AV companies and research. For example, the government established the Singapore Autonomous Vehicle Initiative (SAVI), which oversees and manages AV research, test-bedding, and the development of applications and solutions for Singapore residents by industry partners and stakeholders (nuTonomy’s partnership with the Singapore Land Transport Authority is part of SAVI).
We are optimistic that the government of Singapore will put in place a streamlined and thoughtful approach to permitting large-scale deployment of autonomous vehicle services. nuTonomy’s software has been tested in the US, Singapore, and Europe, and we look forward to further expansion as the popularity of autonomous vehicles grows worldwide.
Google Car (now Waymo) has transitioned from the original vision of building a self-contained system within a specially-built car to developing technology for existing manufacturers. Does this signify a setback or a future trend?
We do not think this is a setback, as this is exactly what nuTonomy is doing. We aim to integrate nuTonomy's software with various vehicle platforms (provided by a vehicle partner) to offer a mobility service. This will look similar a taxi or ride hailing service of today, with the key difference that there will be no driver at the wheel. By removing the driver, we can radically reduce the cost of our transportation service, while increasing safety and efficiency.
What is the common thread between being a research scientist, a writer and an entrepreneur?
These activities share a common thread of creativity, and a requirement to create something truly new – whether it's a scientific article, a work of fiction, or a start-up company.
Jeanne Yizhen Yin
Find more analysis articles & interviews in our 2017 Innovation, Technology & IP Report.