Just last week the Mayor of Las Angeles announced that L.A. will be the first city ready for self-driving cars. This is a bold statement considering that no other city has really taken a major plunge on infrastructure for autonomous vehicles, and thus we have nothing to which we can compare the actions of L.A. The notion of Uber-like services for autonomous vehicles is fairly compelling, but we need to ask - what kind of infrastructure is appropriate or needed for this product-service system? For example, should the police be integrated into an alert system incase things go bad? To take it further, what issues should urban planners consider regarding autonomous vehicles. What does the future city need?
Much of the technology that will shape our urban experience tomorrow is not quite mature enough to meet general expectations - but that is why planning for it today is important. Having the discussion from a planning perspective will reflexively shape the technological development and diffusion, giving us the opportunity to take responsibility for our lives - much like the public health movement of the industrial era. Recently, many car makers have announced plans to roll out self-driving cars in the next few years including Cadillac and BMW. Likewise, Audi has received the first permit in California to road test self-driving vehicles. While MIT's Technology Review last summer, we are nowhere near the point in time when fully autonomous cars will be fully functional, and we are in fact several decades away - so now is the time to discuss how to make this work.
As we shed the remnants of the industrial era and step into the second machine age, let us imagine how it will look. Autonomous vehicles are not limited to cars. Flying aircraft, delivery UAVs, boats, underground tunneling devices, and robots of all kinds can be expected to appear in the next few decades. A primary part of the challenge in creating these technologies is that the urban environment is highly stimulating, directing millions of cues toward a driver at a given second. The sounds of crosswalks, the recognition of a runner nearing the corner, and the flashing lights of a tow-truck vs a police car or ambulance all provide information to a driver. These are also the elements of urbanism that make cities exciting and interesting. Strip the city streets of its life, and yes, we can make cars that will safely drive themselves through stop and go traffic - but is that where we want to live? Dehumanized modernist vision didn't work the second time (Le Cabusier's Machine for Living?), so what kind of infrastructure and planning do we need?
I don't have all the answers, but I think about this question quite a bit. I have a few ideas and hopefully these will inspire others to explore the ideas more deeply. If you have anything to add, I would love to hear from you.
1. Robust GPS system.
At present there are 32 GPS satellites orbiting the earth, at 20,000 KM above sea level. I know little about satellites, but I can't help but wonder is this a sufficient system for constant global demand. So far so good, but can this meet the future demand if you multiple current use by 10 or 100? How might we improve such systems to refine geolocation. Its important to realize that already a great deal of variance occurs within GPS positioning, so while you might have accuracy within a meter in New York City, your GPS points in a rural and low populated landscape might vary as much as 20 meters. Will this be sufficient when your car drives you to work?
In my last post I wrote about the idea of land use planning for drones. Yet the concept of geofencing does not need to be limited to UAV use. It can also be integrated into self-driving cars, water-based robotics, tunneling machines, and any other form of autonomous vehicles. Creating this system isn't hard, but creating a system of standards for the geofencing to work across cities, states, and nations might be more challenging - which leads to the next concept.
3. MIDI for the City.
One thing that has made the internet blossom is the standardization of HTML, APIs, and data structures like JSON to allow developers to freely port one system/tool with another. In a similar fashion, MIDI provides a set of standards for musical softwares and electronic devices to communicate. To my knowledge there is not a set series of standards for electronic device integration at urban scale. Maybe the internet of things will be the solution, but is IPV6 sufficient to address all these billions of objects?
4. Modular Infrastructure
Many cities will generally suffer in the new economy because there is insufficient growth in Commercial, Off The Shelf (COTS) products for smart city creation. Smart Cities cost millions or billions of dollars and are dominated by companies like Cisco, Siemens, and IBM. Rich cities will have the money to optimize and poor cities will not be able to compete. There are a few of us (ahem), working on the design and creation of modular components, dashboards, and sensor networks that can provide municipal plug and play operation - but this market space hasn't taken off yet. This could provide an opportunity for increased safety, reduced costs, and general improvements in urban life quality at scale but more people need to be working on this.
5. Localize Energy Policy
Sufficient energy systems are a constant problem and a major inhibitor for technology diffusion. With the advancement of autonomous systems our energy demands are going to spike. I think some of the more interesting work in this area is within using ocean driven systems (photo at top of page). Yet we can also harness the simple but functional technologies we have today. Many regulatory energy tools exist at the national level, but perhaps city governments need to be more aggressive in local laws. What if every building permit required new construction to include a solar energy component? What if every historic reuse, preservation tax credit, or publicly funded project mandated the integration of passive energy systems? We haven't perfected passive energy, but whatever we have is only effective if policy matures.