Posts tagged with "self-driving cars":

Placeholder Alt Text

URBAN-X’s latest startups bring AI to urban roads, floating cameras to the skies

At URBAN-X’s latest demo day, held at the nARCHITECTS-designed ADO creative hub in Greenpoint, Brooklyn yesterday, the incubator's third batch of cohorts presented technological solutions to urban problems, ranging from a “smart crane” to collaborative retail for small stores. URBAN-X, a startup accelerator and partnership between MINI and Urban Us, takes on up to 10 companies every six months, invests up to $100,000 in each, and connects them with business and design expertise. The most recent group, with nine companies, debuted products and services that were designed to change the way we live in cities, with a focus on the human-centric experience. Qucit (Quantified Cities) is attempting to improve not only urban mobility, but happiness, through artificial intelligence. While other companies have focused on monitoring narrow bands of things such as transit ridership, street usage, bike docking and other urban information, Qucit wants to integrate all of this information vertically into a cohesive model. By aggregating usage data, Qucit has already helped redesign a dangerous roundabout in Paris, and will be bringing its machine learning services to Downtown Brooklyn for a pilot project in early March. Swiftera is approaching similar problems from the air. By using a balloon and floating a camera above what drones can reach, but below satellites, the company is promising high-resolution imagery at specific locations with a short turnaround. By selling actionable geospatial data to planners, developers, architects and municipalities, Swiftera would be able to help monitor traffic and accessibility, as well as things such as roof conditions. Blueprint Power is addressing the disconnect between the energy grid and buildings by creating a market for the surplus energy that buildings are capable of producing. When the grid is stressed, buildings with co-gen plants or solar panels should be able to transfer their extra electricity back to the larger network, benefiting both the building owner as well as the general public and utility companies. This transformation of buildings into “intelligent energy nodes” would ultimately see the buildings’ energy systems automated and managed by an AI system. The complete list of cohorts and their pitch videos can be found here, as well as a video of their evening conference. While most of the group has already begun working with real-world companies, they will also be seeking venture capital funding in the near future. Keep an eye out for URBAN-X’s fourth cohort, which will be announced in May of this year.
Placeholder Alt Text

Gensler and Reebok team up to design the gas stations of the future

What will happen to gas stations once drivers switch over to electric vehicles? Reebok and Gensler are trying to find out, and have teamed up for a “Get Pumped” partnership that imagines repurposing the gas stations of 2030 as community fitness hubs. First announced by Reebok, the initiative imagines a future where automobiles are all electric, and 71 million of the 260 million cars on the road are autonomously driven. Gas stations are usually centrally located and easily accessible, and Get Pumped proposes adapting them into community fitness centers. “This design work with Gensler allows us to imagine a future where there is zero barrier to entry for an opportunity to work out and be healthy,” said Austin Malleolo, head of Reebok fitness facilities. “Consumers may not need gas stations anymore, but instead of wasting them, we’re recycling them, and maximizing the space so that they become places of community.” Gensler and Reebok focused on three station typologies for adaptive reuse: The Network would transform the interstate rest stop into gyms where travelers can recharge their cars as well as their spirits. Described as the “power grid of the future” by Reebok, these charging stations would feature boxing, spinning, Crossfit, running trails and Les Mils. The Oasis model would turn the larger gas stations typically found on local highways into nutrition hubs, offering farm to table restaurants, juice bars, and yoga and meditation hubs. Outside, passerbys could visit the fresh herb garden or run on a rooftop track. The Community Center scheme proposes repurposing the local community gas stations into healthier living stations, where guests can work out, take a quick nutrition class, or shop for healthy food as their car charges. Because these stations are typically smaller but more densely clustered, each converted community center would work in tandem and form a greater network. While Get Pumped is the first step in laying out a potential framework for changing what the “gas station of the future” might look like, it’s worth remembering the challenges involved. Any gas station conversion would be precluded by an intensive amount of soil remediation, as toxic petroleum often soaks into the underlying dirt. Although this type of adaptive reuse project has certainly been done before, the feasibility of doing so on a nation-wide scale would be unprecedented, especially as more and more stations close and are simply torn down. Still, it wouldn’t be the first time that a big-name architect has tried their hand at designing filling stations.
Placeholder Alt Text

A tiny start-up partners with Peterbilt to roll out self-driving big rigs

As of 2015, over 70 percent of all freight transported in the U.S. was moved by truck. That represents a whopping $726 billion in gross revenues from trucking alone, and each year, trucks haul everything from consumer goods to livestock over billions of miles in the United States. All of those numbers are growing—so much so, that according to the American Trucking Associations, the industry is running into a major driver shortage. Long hours, days away from home, and the stress of driving 80,000 pounds at 70 miles per hour is not for everyone, but one company is hoping to make the task easier through automation. Embark, a small startup based in Silicon Valley, is led by a number of engineering school dropouts. Its goal is to develop affordable semi-autonomous semis using neural-net–based deep learning technology. By developing hardware that can be fitted onto existing truck models, and software that learns as it goes, Embark has quickly and cheaply developed some of the most promising autonomous vehicles in the world. “Analyzing terabyte upon terabyte of real-world data, Embark’s DNNs have learned how to see through glare, fog, and darkness on their own,” said Alex Rodrigues, CEO and co-founder of Embark, in a statement that coincided with the introduction of the technology this spring. “We’ve programmed them with a set of rules to help safely navigate most situations, safely learn from the unexpected, and how to apply that experience to new situations going forward.” Rather than try to replace drivers, or redesign the trucks or roads, Embark is focusing on working with what already exists. Collaborating with Texas-based truck manufacturer Peterbilt, Embark is retrofitting the popular 579 semi models with sensors cameras and computers that can read existing roads and take over driving tasks from long-haul drivers. When the trucks must navigate more complex urban settings, the human driver takes back command. This focus on solving the open-road problem, instead of the entire range of driving situations, has streamlined the development process. Currently Embark is one of only three companies permitted to test autonomous 18-wheeler semis on the highways of Nevada (the other two companies being Freightliner and Uber). With the Peterbilt collaboration and a recent announcement of $15 million in additional financing, Embark has become one of the leaders in the race to automate transportation. While Google, Tesla, and a slew of other car companies target the finicky consumer market, Embark has its sights squarely on a market struggling to keep up with demand. With hundreds of billions of dollars at stake, and billions of pounds of freight being moved, it seems only likely that it will be the self-driving truck, not sports car, that we will be seeing on the road sooner rather than later.  
Placeholder Alt Text

How will autonomous vehicles change the way architects think about cities?

City planning operates on decades-long cycles, while infrastructure is typically built out using forecasts that extend current trends. If self-driving vehicles are poised to deliver the revolution in urban transportation that Silicon Valley has been promising, how should urban infrastructure accommodate them? With less parking spots needed, how can designers effectively reclaim this urban space? Anticipating the Driverless City, a recent conference hosted by the AIA New York (AIANY), brought together Uber executives, planners, architects, and policymakers in pursuit of a holistic approach to adapting to life with autonomous vehicles. Speakers acknowledged the same general themes over and over again, despite their differing backgrounds. With self-driving cars possibly arriving in New York City by early 2018 and real-world tests already happening in other cities, one of the most discussed topics was the need to plan for an autonomous future as soon as possible. Nico Larco, co-director of the Sustainable Cities Initiative, stressed that "planners think in 30-year increments, and autonomous vehicles are already hitting the streets today. Urban planners should be terrified." Autonomous vehicles will touch on every facet of urban life, from water management through the reduction of impermeable roads, to electrical grid infrastructure, and drastically reshape the economy. Larco, and many others throughout the event spoke of the need for government to begin working with planners and policymakers to redesign cities from the ground-up. Leaning on a "people, places, policy" framework is a good starting point, as architects and planners can strategize about how autonomous vehicles could possibly affect each of the three. Sam Schwartz, former NYC Traffic Commissioner and founder of transit planning firm Sam Schwartz Engineering, described how a future society with self-driving cars could tilt towards "good," "bad," or "ugly" outcomes. The ideal scenario would be one where the use of autonomous vehicles has encouraged mass transportation use, acting to move commuters to and from high-capacity transit corridors. Because self-driving cars can pack tighter and don’t need to park, streets would be narrowed and the extra space converted to public parkland. Conversely, in a world where autonomous vehicles are owned only by individuals, pedestrians might be walled off from the street, and our roads might be more packed than ever. According to Jeff Tumlin, principal and director of strategy at Nelson/Nygaard, the way we think about self-driving cars directly stems from concepts first presented at the 1939 World’s Fair. Nearly 80 years later, architects and planners wanting to design for a future with self-driving cars, busses, and trains, will need to go beyond simply extending our current car culture.
Placeholder Alt Text

Las Vegas will test the first self-driving public shuttle in the U.S.

Yet another unusual attraction has joined Las Vegas’ iconic downtown strip. ARMA, an autonomous, all-electric shuttle bus designed by French start-up, Navya, has been introduced to the city's downtown traffic. The pilot project, made possible through Las Vegas’ partnership with the mass-transit company Keolis and the American Automobile Association(AAA), marks the country's first autonomous shuttle to operate freely within real-world traffic conditions. The bean-shaped vehicle carries eight passengers and is equipped with a computer monitoring system, GPS technology, electronic curb sensors, laser sensors, vehicle-to-infrastructure guidance and a variety of camera systems. Constant wireless communication between the vehicle and sensors set up throughout the area provide real-time information for ARMA to navigate its way through the bustling Las Vegas streetscape. While there are no steering wheels or brakes in the shuttle, an emergency stop button and a human attendant are stationed in the vehicle in case of emergency. Through the course of this year-long pilot project, the shuttle is offering free rides on a half-mile loop in the downtown Fremont East Innovation District of Las Vegas. The half-mile route has just three stops – located on Fremont Street and Carson Street between Las Vegas Boulevard and 8th Street – but is to-date the longest self-driving pilot project to operate in a fully-integrated manner within a real-world city environment. ARMA can operate up to 8 hours on a single charge, and reach a speed of 28 miles per hour. Over the next 12 months, the project aims to provide 250,000 free rides and help soothe public skepticism towards autonomous vehicles. In addition to familiarizing the public to these new technologies, this project will provide real-life research on the relationship between autonomous vehicles and pedestrians, bicycles and other human-operated traffic. Las Vegas will serve as a valuable test site for the future of autonomous transport within the urban environment. If Las Vegas can make it work, we can perhaps expect the expansion of self-driving technology on a broader scale throughout the country. Navya expects electric shuttles to be more affordable to maintain than combustion-powered vehicles. However, the cost of such projects are not cheap to install. Currently, a single Navya shuttle is estimated at around $260,000.
Placeholder Alt Text

Michigan’s Mcity selects startups to test self-driving technologies

A tech haven located on the northern campus of the University of Michigan is redefining Michigan's ‘Motor King’ reputation. Mcity is a 32–acre complex designed to mimic urban and suburban city environments. Complete with painted building facades, dummy pedestrians, bike lanes, roads and highway ramps, the controlled laboratory environment eliminates real-world risks and serves as a unique testing ground for vehicle and urban transportation technology. The combination of physical and virtual alteration possibilities within this ‘fake-city’ allow both for current real-life simulations as well as testing for speculative future mobility scenarios. The ability to replicate human-scale urban and suburban environments is vital for conducting tests to enhance current road safety and to plan for our evolving urban future. The facility is involved in numerous research projects, including testing data collection and management systems, studying interactions between motor vehicles and bicyclists, enhancing pedestrian detection and avoidance technology, and improving intelligent parking guidance system. Having access to a state-of-the-art testing facility such as Mcity provides tech companies with unparalleled development opportunities in their work and research. This fall, five emerging startups have been selected to work at Mcity alongside students at the University of Michigan's TechLab. Tome, based in Detroit, works on enhancing bicycle-to-vehicle (B2V) communication within the urban sphere. CARMERA, based in New York and Seattle, are experts in street-level intelligence focused on creating real-time 3-D maps and scene reconstructions vital for autonomous vehicle performance. RightHook, from San Jose, California, specializes in safely simulating harsh conditions to test the resiliency and performance of automated vehicles. Zendrive, of San Francisco, California, aims to increase driver safety through smartphone data collection. PolySync, from Portland, Oregon, builds software infrastructure and tools to develop autonomous vehicle functions.
Placeholder Alt Text

How sensing technologies can reshape architecture, public health, and cities

Carlo Ratti is the founder of the Turin, Italy-based firm Carlo Ratti Associati (CRA) and director of MIT’s Senseable City Lab. In both roles, he explores how technology can improve the built environment and, it follows, our lives. Recently in Turin, CRA completed the Agnelli Foundation headquarters, which employs a smartphone app to let occupants set personal temperature preferences that the building translates into individualized “thermal bubbles.” Meanwhile, the Senseable City Lab has been a steady source of data and visualization projects—such as mapping walking, running, and cycling trips in Boston and San Francisco—while dabbling in related fields like robotics. AN talked to Ratti about the opportunities and risks that wired, sensing, and smart architecture will bring.

The Architect’s Newspaper: How do you feel about the term “smart cities”? It has become very loosely defined and can refer to anything from superfast fiber-optic networks to autonomous vehicles.

Carlo Ratti: To be frank, I don’t feel that great about it. As you say, “smart city” is often used in a loose way. Also, too many times it equates to top-down approaches in the implementation of urban technological solutions—à la Masdar or Songdo. Our vision is different. When we started the Senseable City Lab at MIT and our design office Carlo Ratti Associati around ten years ago, we were interested in how our cities could become more “sense-able”: able-to-sense, sensible, and perhaps even more “sensitive.” And this has remained our main focus since then.

In many cases, that “sensing” means collecting masses of data—whether it’s trees, human movement, or ride-sharing potential—to reveal new efficiencies, solutions, or patterns.

We need to go back to the very notion of design. According to Herbert Simon, “the natural sciences are concerned with how things are…design, on the other hand, is concerned with how things ‘ought to be.’” I believe that designers must challenge what exists today, introduce new and alternate possibilities, and ultimately pave the way toward a desirable future. In this process, it is vital to get people’s input, which often happens online. In our latest book (The City of Tomorrow, Yale University Press, 2016), we call this method futurecraft.

But do you think there’s a privacy risk that comes with getting people’s input—their data? Are you concerned about surveillance?

I am very concerned—but more about what is happening in our pockets than about what is happening in our cities. Most of our activities—where we go, how we get there, what we buy, to name just a few—are recorded thousands of times every day and stored somewhere in the cloud. Who controls the data? How can we avoid data monopolies? Such questions are more topical than ever.

I would rather prefer a society where value comes not from data itself, but from what intelligence can extract out of it.

That intelligence can really benefit a society. Your projects Underworlds and Urban Exposures explore the public-health benefits of technology—something that is often overlooked in this discussion.

Both projects focus on data to provide a better understanding of human health in the city—which in turn can inspire policy action. Urban Exposures, for example, combines data from air quality measurement stations and human mobility to estimate human exposure to pollutants in a more accurate way.

Currently, it seems European cities are ahead of American cities when it comes to using technology for the public good.

It’s hard to generalize. Europe is very heterogeneous—Copenhagen or Stockholm are very different than, say, Valencia or Athens. What I often notice in the U.S. is a bias against government spending in public infrastructure—perhaps a soft version of “The Plot Against Trains” described by Adam Gopnik in The New Yorker. At the same time, we will find out soon how the trillion-dollar plan by President Trump on U.S. infrastructure will be spent—hopefully not just on walls...

On that topic, new technologies are creating huge opportunity to change—and profit from—how we move within existing infrastructure. As carmakers and ride-sharing companies race to capture that market, would you like to speculate on what “mobility” will look like in 20, even 50, years?

Autonomous vehicles promise to have a dramatic impact on urban life, blurring the distinction between private and public modes of transportation. “Your” car could give you a lift to work in the morning and then, rather than sitting idle in a parking lot, give a lift to someone else in your family—or, for that matter, to anyone. As a result, a single vehicle can go from one-hour usage per day to 24 hours. Under such conditions—and increased sharing of rides—we have calculated that the mobility demand of a city could be met with just a fraction of today’s vehicles.

There could also be dystopian scenarios, however. Car transportation could become so cheap that it might drain customers from subways and buses, turning our streets into an instantaneous gridlock. The impact of autonomous vehicles will depend on the policy decisions we make. I agree with my friend Robin Chase, founder of Zipcar, when she says that “simply eliminating the drivers from cars, and keeping everything else about our system the same, will be a disaster.”

Beyond urban infrastructure, your firm is focusing on infrastructure at the building scale—interior climate systems specifically. This goes all the way back to a 2011 paper you helped author, but most recently it was the focus of your app-driven design for the Agnelli Foundation headquarters. Why this interest in temperature, occupancy, light, and energy usage?

The 2011 paper shows that a staggering amount of energy is wasted on heating empty offices, homes, and partially occupied buildings. That finding prompted a series of projects where we tried to better synchronize human presence and climate control. The Agnelli Foundation, which opened just a few months ago, is the first application of such ideas at the architectural scale. We equipped a historical office building with digital sensors that monitor variables such as temperature, lighting levels, and matched this data with occupancy information.

When a person gets into a building and sets her preferences in term of temperature or lighting, the building-management system recognizes her and automatically responds by activating the system accordingly. It generates something like a “thermal bubble,” following a person within the building as she moves across the various rooms and corridors. The final outcome is better comfort for users as well as a substantial reduction in energy consumptions—estimated up to 40 percent. When all occupants leave, the room returns naturally to “standby mode” and saves energy, much as a computer would do.

We imagine that more and more buildings will be equipped with sensor networks—making architecture increasingly able to “sense and respond.” By designing climates, we might get closer to the vision of architecture as a third skin—an endlessly reconfigurable space able to adapt to human needs.

Placeholder Alt Text

A Michigan facility is the nation’s epicenter for testing self-driving cars

The race is on to develop connected and automated vehicles (CAV) that are viable and affordable. The road to this goal is not a simple one, though. While Silicon Valley is working on the software side of the challenge, the U.S. government is looking back to the place where it all began: Michigan. Specifically, Willow Run in Ypsilanti Township.

Willow Run was a B-52 manufacturing plant during World War II. Today, the site is in the middle of a transition that will make it the epicenter of automated-vehicle research. Willow Run is now home to the American Center for Mobility (ACM), and it has been designated as the first national CAV proving grounds by the Department of Transportation.

At over 500 acres, the center includes a variety of environments designed to simulate real-world situations. These include a 2.5-mile highway loop, a 700-foot-long curved tunnel, two double overpasses, and multiple intersections and roundabouts. Matched with Michigan’s varied and sometimes extreme weather, the center provides everything needed to put new autonomous testing technologies through their paces.

The first task of the facility will be to help establish voluntary standards for CAVs, infrastructure, and autonomous technologies. Along with the Institute of Transportation Engineers (ITE) and SAE International, the center will work to identify the most immediately needed guidelines for safe automated transportation. The center is also working with the University of Michigan’s Mcity, a smaller research facility with its own proving ground. Mcity’s position within the greater university allows researchers access to the school’s engineers, public policy experts, and law, business, social sciences, and urban planning faculty.

As a public-private partnership, the center is also working with companies like Toyota and AT&T. Toyota, which already does automated research at Mcity, recently invested $5 million into the center. AT&T is providing a dedicated LTE cellular network needed for the communication side of the CAV equation.

“As we move forward with the development of autonomous cars, we must remember that not all test miles are created equal,” said Gill Pratt, CEO of Toyota Research Institute, at the announcement of the Toyota-ACM collaboration. “The road to creating a car as safe, or safer, than a human driver will require billions of test miles including simulation, real-world driving on public roads, and closed-course testing where we can expose our systems to extreme circumstances and conditions. The new ACM closed-course facility is a significant step forward in this journey and will accelerate our ability to help prevent crashes and save lives.”

According to the World Economic Forum, 10 percent of vehicles in the U.S. will be driverless by 2026. Before that can happen, new hardware and software will have to be developed to overcome issues of trust, cost, efficiency, and safety. The U.S. government is counting on Michigan’s automotive brain trust to solve these issues and move the country back into the lead position in the automotive industry.

Placeholder Alt Text

Will the rise of self-driving vehicles signal the death of the traffic light?

The dawn of self driving cars promises to be an exciting new era for transport. However, what exactly lies ahead is still up for debate. Researchers at the Massachusetts Institute of Technology (MIT), the Swiss Institute of Technology (ETHZ), and the Italian National Research Council (CNR) have outlined how traffic signals could be rendered obsolete if automated vehicles get their way. The development is known as "slot-based intersections," and if realized, would significantly reduce queuing, delays, and pollution. If evidence from any science fiction movie is anything to go by, it's that humans have very little trust in automated technology. It's easy to picture: panic as your self-driving car appears to be careering into another, only to miss by a a tiny margin, all perfectly predicted by an automated system of course. https://vimeo.com/106226560 That may be an exaggeration, but Professor Carlo Ratti, Director of the MIT Senseable City Lab and his team have produced a model that shows cars zipping through a four-way intersection both without stopping or slowing down and remaining unscathed. “Traffic intersections are particularly complex spaces, because you have two flows of traffic competing for the same piece of real estate,” he said in a press release regarding the study, published in detail here. “But a slot-based system moves the focus from the traffic level to the vehicle level. Ultimately, it’s a much more efficient system, because vehicles will get to an intersection exactly when there is a slot available to them.” https://youtu.be/4CZc3erc_l4 Trust in such a system would have to be high. Communication between cars would have to be flawless and safety measures for failure would also have to be in place. That said, if implemented, the system would speed up journey time and also reduce pollution by cutting down on the time spent idle at traffic signals. Of course, signal-less interchanges already exist, they're called roundabouts. But the possibility for human error (and hence collisions) still exists in the roundabout, along with the need to give way to others.
"Slot-based intersections are similar to slot-based management systems used for air-traffic control," say the team. "Upon approaching an intersection, a vehicle automatically contacts a traffic management system to request access. Each self-driving vehicle is then assigned an individualized time or “slot” to enter the intersection." Speed limits could also change. If a perfect system can plot every movement, why not travel at the fastest, yet safest, possible speed? This is just one of the questions arising as self-driving cars become more and more likely to enter our lives. Would car lanes also be made thinner? Vehicles won't be making mistakes so why not cram as many in as we can and maximize efficiency? https://youtu.be/sQuJ8GKTjFM In terms of having a central traffic organizing system, getting different car manufacturers to be completely open with each other is another major bridge that would need to be crossed. And as for the more pressing issue of automated vehicles' interaction with humans, MIT's Senseable City Lab responds by saying: "slot-based intersections are flexible and can easily accommodate pedestrian and bicycle crossing with vehicular traffic."