Event Attendance: Designing the Connected City @ Cooper Hewitt, Smithsonian Design Museum

By: Michelle Kung
INFO 601-02 Assignment 3 Event Attendance

Cities like New York are notorious for congestion and pollution. It often takes the same amount of time to walk somewhere as it does to drive somewhere. But big tech companies are reimagining urban mobility with connected and autonomous vehicles (AVs). On the 25th of February, 2019, leaders in the field of autonomous vehicles or driverless cars came together at the Cooper Hewitt, Smithsonian Design Museum for a panel discussion. Moderated by Cynthia E Smith, the curator of Socially Responsible Design, the panel consisted of Sarah Williams, the director of Civic Data Design Lab at MIT, Ryan Powell, the head of user research and UX design at Waymo (Google’s self-driving car project), and Jack Robbins, the director of urban design at FXCollaborative. With diverse backgrounds, the three panellists debated topical issues engendered by AVs.

A World Unknown

One thing that the three panellists agreed on was that no one really knows how technology will impact mobility in urban spaces. The field is still new and concepts have only been tested on limited scales. Jack Robbins called this ‘a new era of mobility’. Indeed, we have no idea how the way we move, not just within cities but across the country, is going to change. All we know, and all leaders in the field know, is that autonomous vehicles will be the biggest drivers of change.

Heaven or Hell Scenario

Jack Robbins illustrated two opposing scenarios: a heaven scenario and a hell scenario. In the heaven scenario, after autonomous vehicles have replaced standard vehicles. There will be fewer vehicles on the road, fewer vehicle miles travelled and more spaces freed up in cities for other things. Without the need for parking (i.e. temporary storage of private vehicles) within the city, there is a tremendous opportunity for the creation of more green spaces and open spaces which will increase the liveability of any congested and densely populated city. On the other hand, in the hell scenario, there will be more vehicles and more vehicle miles travelled. Autonomous vehicles will be on the road driving around with or without passengers, which would be terrible for inhabitants of cities as well as the health of the planet. 

Will companies deliver on their promises?

Companies developing autonomous vehicles are of course promising everything detailed in the heaven scenario. But Jack Robbins cautioned event goers against trusting these companies too much.  After all, the way they make their money is incompatible with the promises they are making. For example, Google sells advertising but is promising increased road safety, mobility equity, easy parking, transit support, and less traffic. But how? By gathering an increasing variety of information on humans and on built environments.

Human Behaviour is Information as Thing

Waymo, Google’s driverless car company, purports to take a human centred approach to create a ride hailing service. Their primary goal is physical safety. In order to achieve this, Waymo collects an incredible amount of data on people and human behaviour in order to program the world’s most experienced drivers. According to Ryan Powell, 94% of road accidents are caused by human errors and Waymo’s aim is to eliminate this altogether. Waymo has managed to collect the data of behaviour patterns of adults, children, and cyclists in order to teach their fleet of driverless cars how to react safely in each scenario.

On the surface, treating human behaviour as information as thing is not at all revolutionary. Psychology, anthropology, and a whole host of other social sciences have studied the behaviour of humans for decades. But the monetisation and capitalisation of this information on such a large scale is new. Speakers in this talk were more interested in talking about the information regarding the space and infrastructure of a city than information about the people living in them, which is slightly alarming.

Public Space as Private information

A huge topic of debate in this design talk was the importance of the public nature of public space. Speakers Sarah Williams and Jack Robbins both challenged Ryan Powell on Waymo’s current practices of keeping information about the public space private.

As Waymo gathers more and more information on public spaces, their data set becomes more valuable. Sarah Williams advocated for city governments to leverage their power to ban companies like Waymo from operating in their cities to negotiate data rights. Both Sarah Williams and Jack Robbins argued for the importance of public governing bodies to step up and play a more active role in this sphere rather than passively hoping for technology companies to do the right thing by citizens. Autonomous vehicles pose real dangers in deepening and widening the digital divide, privatising public data, and decreasing equitability in cities. It is up to cities to set boundaries, guidelines, and regulations so that data collection and ownership of cities contribute to the public good and can benefit the many rather than the few.

Conclusion and Reflection

This design talk was fascinating and helped me conceptualise the new forms of information this emergent technology creates. The panel discussion really encouraged me to think more deeply about the data rights of citizens and city governments. It is already inconceivable, the amount of data big software companies have on our digital behaviour. It is entirely unimaginable, for the average user, what information companies developing autonomous vehicles will have on our behaviour in physical environments once AVs become more mainstream.

In the meantime, it is clear that city governments need to catch up to big tech players in order to ensure that public spaces are protected, new infrastructure built is adaptable to unforeseeable changes, that cities become more liveable in the long term for all its inhabitants, not just a select few.


Buckland, Michael K. “Information as Thing.” Journal of the American Society for Information Science42, no. 5 (1991): 351-60. doi:10.1002/(sici)1097-4571(199106)42:53.0.co;2-3.