TEDx full Transcript


A beautiful orchestration of code, lasers, and machine learning, driverless vehicles offer us the promise of harmonizing our congested streets while also making them safer, rescuing us from our own error and fallibility.  Software is being written so that, at the tap of a button, a self-driving car will know where to pick up and deliver your groceries, when it’s time to get you to work, or at which club to come fetch you and your friends after a wild Saturday night. Paying for parking will be a thing of the past, so will parking tickets, not to mention all that time wasted circling to find the perfect spot. As they learn how we to move around, these cars gather data and create beautiful 3d maps of our world.  

Ladies and gentlemen, we are at the frontier of a new Mobility Revolution.  Code is the new concrete. Lasers are the new eyes on the road.  Data is the new fuel. And with that, we are told that driverless vehicles have the potential to save lives, improve traffic congestion, and cut down on emissions.

They better.


In 2013, I coordinated a 33-mile driverless car ride for the chair of the house transportation committee and the then-secretary of PennDOT. Since that time, two of my dearest friends have been killed by cars.

In December of that year Andy, a doctor exercising the hippocratic oath after a 50 car pile up, stepped out of the car where his wife and two children were to tend to the victims of the crash, only to be struck and killed, himself. Susan on the other hand, was simply was waiting on her bike at a stop light on her way home from work.  When someone who shouldn’t been behind the wheel pinned her between to cars causing her to lose her life immediately.

I’m not alone in my grief. 40,000 people die in roadway crashes every year in the US. 1.3 million people every year around the world.

I felt intense frustration continuing  my work at the United States Department of Transportation’s  National Research Center for Transportation Safety. As an expert in this field, Autonomous vehicles give me hope.  We need them to fulfill their promise. But I want them to work for all of us and I’m not convinced we even know what that means yet.

First let’s talk about those parking tickets. NYC raked in $545 million in parking ticket revenue in 2017. Parking tickets are annoying at best but they fund our cities.  As an individual, paying $30 sucks, but on the city level it can help maintain our streets. The downward trend we are seeing in car ownership right now coupled with cars that can drive themselves means the need for parking, especially on-street parking, dramatically drops.

In other words, the way we fund infrastructure in this country is about to be toast.  

So, if the cars aren’t parking, what are they doing?  They are roaming, moving around all around on the street all the time.  Waiting for someone to call them. Why is this bad? Emissions, congestion, and safety.

But what about those cars making beautiful maps of the world? Their goal is not just to move us around. It’s to understand our every move.  

And we just hand our data and privacy over to them.  

What could we, the people, be doing now to make sure we reap the benefits without these harms?

Right now, what has become so very clear to me is that while the best and the brightest in the world are designing gorgeous algorithms, we are missing a major piece of the equation - humanity.

What if we, the people, demand that this new mobility also be orchestrated with our values, tailored to our community’s needs, like affordable housing and economic access?

What if this mobility revolution could guarantee that data will be balanced with privacy, that implicit bias won’t be written into code.

In my world, I hear over and over again from cities that they want data from the driverless and ride-sharing car companies.  They just need to know where people are coming from and going so they can optimize transit stops and utilize this data for other city purposes.

In the past we let cars lead the design of our roads, which have cut off communities, created inequities and driven suburbanization, we have to understand, now, that if we let technology drive this revolution, we, the people, will lose.

So here is our opportunity to engage.

We, have  a lot of data that we are just giving away. We are letting companies develop their products on our roads, in the public right away, with us as the crash test dummies,.

What if we flipped this concept on its head?   The companies should be expected to share what they are learning with we, the people.  What if we could use those 3d maps they are building in real-time to have real-time insights about our own cities?  

It would be more powerful to know where people are NOT going because they can’t get there? What if, we, the people, gather our own data and turn around and license that to the industry? Forming community data trusts, while also providing powerful insights  to our local governments for planning purposes.

Right now cities rely on money from parking, what are they going to do when parking disappears? How are we going to charge the autonomous ride-sharing cars that aren’t parking?  They are using our streets, they are using our resources. Industry and government are talking about charging per mile. How do we empower cities to capture this funding?  What if cities found powerful ways to reuse the land that is freed up from parking to address community needs, like affordable housing infill development or fresh food markets?

How do we make sure we are not creating a river of cars and freight that are constantly roaming around our streets?  What if we we limit them certain streets, specifically with freight, certain times of day. Then if we move them to designated streets, again, we are freeing up land, for we, the people.

So are cities and communities equipped to harness this potential future?

I’m afraid not.

In terms of safety, the Federal Government is moving tentatively in its regulatory process.  Basically, right now it is only voluntary for Driverless testers to submit the steps they have taken to ensure that their machines are safe on our roads.  What if it was mandatory?

States are all over in their approach to governing driverless cars.  From California to Pennsylvania - who has adopted an innovative soft law approach focused on consensus building and collaboration.

So, how do we, the people trust this beautiful orchestration of code and machine learning that is on our streets, testing right now?  Our first automotive safety standards were established in the 60s with the help crash test dummies tests. The same needs to happen for the software that is being incorporated into these cars.  We need to consider policies requiring a standard for testing. That way we can avoid incidents like we saw earlier this year in Arizona where a pedestrian lost her life to a confused machine as she crossed the road.


40,000 people killed a year by human driven vehicles, yet driverless cars have the potential to save so, so many loved ones, like your Susan and your Andy.

And in order to create a future with principles and policies that match our human values,

We, the people need to be in the driver’s seat.