The Driverless Commute: The four policy areas that cities must address to make the autonomous era people-centric; AVs positively last-century bugs; and robo-race car sets world record on speed.

1. Cars ruined the American city. Can AVs save it?

Since at least the 15th century, the era’s dominant mode of transportation has always had a way of exerting itself, often unfavorably, on the fabric of cities, but nowhere responded to the automobile with the same ill-considered enthusiasm as the American city, which forfeited one of its most precious public social spaces—its street—in a way that Europeans never did.

Now, urban planners from the United States’ largest cities believe the arrival of autonomous vehicles represents a rare opportunity for car-subservient downtowns to reset the board finally.

This week the National Association of City Transportation Officials, an 81-city coalition that includes New YorkBostonAtlantaLos Angeles, and Seattleunveiled a 131-page blueprint to refocus contemporary urban planning to take advantage of self-driving cars. 

The plan envisions a future of more parks and fewer parking lots by addressing high-capacity transit, smart-city data collection and use, congestion pricing mechanisms, and the delivery of urban freight.

Read More

The Driverless Commute: GM, like Waymo before it, missed an AV deadline, but what does it mean for the industry going forward?; tension building between AV, transit advocates

1. The big question before AV makers: first, or best?

General Motors Cruise Vehicles

As the unit economics of driverless cars continue to spiral, the industry has been forced to soul-search its group psychology: move fast and break things because if you ain’t first, you’re last.

But with consumer confidence in the technology (and the people developing it) slipping to historic lows, technologists and carmakers are sobering to the reality that half-baked deployment could kill the nascent industry in utero.

The root: Validating safety and performance in diverse, chaotic and multi-hazard environments has proven more challenging than engineers first believed. Not to mention, costlier.

Read More

The Driverless Commute: VW and Ford to partner on AVs, latest in long string of tie-ups; 11 companies unveil safety-as-design principles, offering closest thing to industry standard; and Lyft tests its cars on blind passengers

1. Big tabs and hard realities.

Going-it-alone is, like, so 2018.

Volkswagen and Ford, one-time rivals fast sobering to the costs and difficulty of engineering next-generation cars, said Friday they would pool resources in the development of autonomous vehicles. Under the long-rumored deal, VW will invest upwards of $2.6 billion into Ford’s self-driving unit, which was already valued at $7 billion before the tie-up.

The agreement is the latest in a string—so many, in fact, that we’ve lost count—of fiercely competitive carmakers cooperating to develop self-driving technology.

Read More

The Driverless Commute: Apple stuns with Drive.AI acquisition; did Florida go too far in new AV bill?; Toyota throws in with Baidu’s Apollo project; and Waymo rolls out limited partnership with Lyft in Arizona

1. Assumptions and expectations

Apple, the famously secretive consumer electronics giant, said this week it had acquired self-driving startup Drive.AI, whose human-robot interaction systems and deep-learning approach earned it an outsize reputation in the autonomous constellation.

The days of going-it-alone are behind us.

  • Like many struggling to reconcile real-world deployment challenges (it turns out, engineering self-driving cars is a lot harder than marketers promised) with stratospheric expectations, Drive.AI had come into hard times recently. According to reports, it filed paperwork ahead of the Apple announcement that it intended to dissolve and lay off its entire workforce.
  • Previously, the company had a variety of splashy pilots under its belt, including a recent test in Texas in which human contingency drivers had been removed from some vehicles.
Read More

The Driverless Commute: Is 2019 a bear turn for AVs?; the fight over connected car communication divides car makers in Europe; cheaper, lighter LiDAR and machine-learning for lane-keeping.

1. A bear year

2018 was the year of the driverless car.

No, they didn’t become commercially available and they failed to traverse the realm of true autonomy, but they captured the public’s imagination and sometimes paranoia in a way unlike ever before. Finally, it wasn’t just artificial intelligence researchers talking about autonomous vehicles, but regular Joes drawn in equal parts fascination and fear to the subject.

But now the honeymoon is over.

After throwing money at startups like drunken sailors for years, the industry has signaled a coming retrenchment. Rather than more multi-billion dollar acquisitions, expect to hear about reoganizations and partnerships that would have been unthinkable only years earlier.

Read More

The Driverless Commute, presented by Dentons: How programmer bias will be reflected in AI platforms; a case for public investment in AV access; and Waymo, GM & Ford top new AV leaderboard

1. When an engineer’s implicit bias becomes a computer’s

Autonomous Vehicle (AV) LIDAR sensor

Can the machine learning algorithms that underpin the operation of autonomous vehicles perpetuate—or worsen, even—social, structural biases against people of color?

Maybe, according to a new paper (PDF) from the Georgia Institute of Technology, where researchers tested the accuracy of object detection systems (not unlike those used in driverless cars) in positively identifying pedestrians of varying skin colors.

Researchers paid human test subjects to review a collection of 3,500 images of people of varying skin tones and to mark each photograph with “LS” or “DS” to designate light or dark skin of the subject.

Read More