I previously blogged about Volkswagen’s “fun theory” initiative, which is about using fun to encourage proenvironmental behavior (e.g., recycling, taking the stairs vs. the escalator). My adviser, James Landay, sent me this environmental group’s appeal to Volkswagen to improve the fuel efficiency in their vehicles (image below). Perhaps we need to embed some “fun theory” into the CAD software that VW engineers use to build their vehicles or the spreadsheets VW business executives use to count their profits. The image was created by Friends of The Earth (click on the image to maximize).
The Bay Bridge in San Francisco was closed for nearly a week due to the collapse of a steel beam and two tie rods. During the closure, the ridership on the Bay Area Rapid Transit (BART) system increased dramatically to record levels. So, now BART is studying ridership data and feedback from new and infrequent riders, in hopes of attracting them to take public transit on a regular basis.
On Wednesday, the first full day of the emergency bridge closure, BART began an online survey aimed at finding out more about those reasons. The survey will close at the end of business Tuesday, Nov. 3, so if you used BART during the bridge closure, there’s still time to submit your feedback.
Around 1,500 people responded to the survey, which was posted on the homepage of BART’s website and promoted through social web channels including @SFBART on Twitter, the SFBART blog and Facebook fan page. Although anyone could take the survey, analysis will focus on the responses from first-time or infrequent riders.
Preliminary Results
Suggestions given in verbatim, open-ended comments for what would get people to ride BART more frequently included: expanding service, improving parking availability at stations, making machines easier to use, ensuring announcements and signage are clear, keeping trains clean and providing more police presence. BART will dig deeper into the statistical data from questions about trip origins, destinations and frequency.
Carbon Savings
Getting more people out of their cars and onto trains is good not only for BART, but also for reducing environmental impacts of highway congestion, he said. For example, during the first two full days of the bridge closure on Wednesday and Thursday, BART estimated that riders took 163,000 extra BART trips. If they had driven vehicles for those trips, the trips would have resulted in about 1.8 million pounds of CO2 emissions.
Technology has been cited as one possible solution for increasing ridership amongst choice riders (choice riders are those that have multiple alternatives to travel). Studies of the OneBusAway system, for example, have shown that real-time information about city bus arrivals/departures, can increase the number of rides that people take (though it’s yet unclear whether OneBusAway increases rides amongst choice riders). The BART webpage also links to a variety of iPhone and other mobile phone apps for the BART.
On Friday, October 9th, I was part of an invited panel at the Walk21 conference on Using Powerful Web Apps to Build a Livable Streets Movement hosted by Nick Grossman from The Open Planning Project (TOPP) Labs. Other panelists included Ben Berkowitz from SeeClickFix, a tool to report and monitor community issues; Aaron Ogle from WalkShed.org, a visualization tool to explore very precise and personal walkability calculations; and Seth Priebatsch from SCVNGR, a website to host geo-based scavenger hunt games. It ended up being a tremendously successful panel with a very fruitful discussion which included questions about privacy, the pros/cons of transparency, motivating adoption, and government engagement. Discussions will continue on the mailing list: streets-advocacy-tech@googlegroups.com.
The title of my talk was The Feetback Cycle: Leveraging Everyday Technologies to Change the Way We Move. I focused on the emerging area of Persuasive Technology and the ways in which technology may be used to encourage particular behaviors. I began the talk with a brief overview of popular behavior motivation techniques, highlighted past studies by Sunny Consolvo and colleagues at Intel Research exploring the use of mobile phones to promote fitness activity and then transitioned into a lengthier overview of the UbiGreen Transportation Display. Unfortunately, due to time constraints, I was not able to go over commercial offerings of persuasive technology like the Nike+iPod, the newly released iPod Nano Pedometer or the long-awaited FitBit but you can see the slides here (pptx file, 33.9 MB).
Below are some pictures from the talk itself:
The Toyota Prius is perhaps the quintessential eco-feedback system, it provides real-time information about a driver’s fuel efficiency as well a historical graph to track progress over time.
Back in 2005-2006, Sunny Consolvo and colleagues from Intel Research, Seattle used a pedometer and mobile phone to show that rewards mediated by a technology could be effective in motivating fit behavior even if that reward was simple. In this case, study participants were rewarded with an asterisks when they achieved their step goals.
The UbiGreen Transportation Display semi-automatically senses transportation modes such as bicycling, running, and walking and feeds this information back to the user with the goal of motivating green transportation decisions.
The UbiGreen Transportation Display uses the background of the mobile phone (sometimes called the wallpaper) to display evocative imagery that changes based on sensed transit activity (sort of like a real-life Choose Your Own Adventure where the choices are sensed in the physical world rather than in a book).
I just got done presenting at IJCAI09 on the shared bicycling research I conducted while a visiting researcher in the summer of 2008 at Telefonica Research in Barcelona, Spain. This is joint work with Joachim Neumann and Nuria Oliver (both of Telefonica Research). You can download the talk (PowerPoint slides) here.
Abstract
City-wide urban infrastructures are increasingly reliant on network technology to improve and ex-pand their services. As a side effect of this digitali-zation, large amounts of data can be sensed and analyzed to uncover patterns of human behavior. In this paper, we focus on the digital footprints from one type of emerging urban infrastructure: shared bicycling systems. We provide a spatiotemporal analysis of 13 weeks of bicycle station usage from Barcelona’s shared bicycling system, called Bicing. We apply clustering techniques to identify shared behaviors across stations and show how these behaviors relate to location, neighborhood, and time of day. We then compare experimental results from four predictive models of near-term station usage. Finally, we analyze the impact of factors such as time of day and station activity in the prediction capabilities of the algorithms.
Some pictures (with captions) from the talk:
Our focus was on utilizing existing urban infrastructure to sense data about human behavior that is *freely* available (e.g., not proprietary data but data that we can freely access). In this case, we use shared bicycling usage to uncover spatiotemporal patterns of human mobility in the city of Barcelona.
We have reached a pivotal point in time where city infrastructures are transitioning from mechanical/analog systems to digital systems thereby creating digital traces of human activity. Bruno Latour notes the potential to access the masses of data that are of the same order of magnitude as that of the natural sciences.
Our main contributions were: (1) demonstrating the potential of using shared bicycling as a data source to gain insights into city dynamics and aggregated human be-havior; (2) exploring the relationship between spatiotemporal patterns of bicycle usage and underlying city behavior and geography; and (3) studying patterns in bicycle station usage, including the prediction of usage patterns and an analysis of how factors such as the time of the day affect this prediction.
We obtained our data by scraping the bicing website once every two minutes. We downloaded station geolocation information as well as the number of current free parking spots and number of currently available bicycles.
One of our motivations to explore prediction was the fact that 66% respondents to an online survey about Bicing stated that they had difficulty finding a free parking slot when trying to drop off a bicycle. This is a major impediment to Barcelona residents adopting Bicing as a primary form of transportation as searching for a station with a free parking spot takes time. Indeed, 50% of respondents avoid Bicing when they are traveling to a place where they must be on time.
We used dendrogram clustering on station temporal usage data to see how Bicing usage patterns are shared across the city. We also explored how our prediction algorithms performed in relation to these clusters.
One longterm goal of our work is to explore ways to make shared bicycling more self-sustainable. Current shared bicycling systems rely on trucks to load balance the bicycles (i.e., to make sure they are well distributed throughout the city). We are looking at ways to incentivize bicing users to drop off/pick up bicycles slightly out of their way to reduce the maintenance/operating overhead on the city. A mobile phone application could recommend a station close to a user’s final destination that is predicted to have a need for bicycles.
This work would not have been possible without my colleagues Joachim Neumann and Nuria Oliver. Joachim, in particular, worked tirelessly on this project for six months and was absolutely essential to many parts of the project including data logging, model building, and evaluation.
A team member from our HydroSense project sent along a link to this article, Start-ups are racing to get electric motorbikes to market, in the LA Times. He has a contact at Mission Motors, who is featured in the article. Mission Motors just recently unveiled a prototype (pictured above) for its 150-mph, 150-mile-range electric motorcycle at the TED conference in Long Beach last month.
Highlights from the LA Times article:
The Electric Motorcycle Scene
Since 2007, when Vectrix of Middletown, R.I., first rode onto the scene with its battery-powered Maxi Scooter, a growing number of U.S. start-ups have entered the plug-in two-wheeler market. They’ve invested millions of dollars in vehicles, many of which are poised for production within a year. Honda and Yamaha have said they’ll be coming out with electric motorcycles in two years. Though a rapidly deteriorating global economy and relatively low gasoline prices may not seem like ideal conditions for launching or ramping up a company in an unproven field, many of the two-wheeled-EV start-ups say they have benefited from it. Craig Bramscher of Brammo Motorsports in Ashland, Ore., says he raised $10 million in venture capital last year — all of it after the financial system froze up in September.
Why Electricity?
“It’s amazing how inefficient the vehicles we’re driving today really are,” said Forrest North, founder and chief executive of Mission Motor Co., a San Francisco company that unveiled the prototype for its 150-mph, 150-mile-range electric motorcycle at the Technology, Entertainment, Design conference in Long Beach last month. “Electricity is just so many orders of magnitude more efficient that it’s the only way to go,” said North, a former mechanical designer for Tesla and leader of Stanford University’s solar car team in the mid-1990s.
Like many EV entrepreneurs, North, 33, had looked into hydrogen and biodiesel as power sources but found them impractical. Hydrogen is abundant, but turning it into fuel and developing a distribution infrastructure are costly. Biodiesel can take more energy to produce than it generates.
With electricity, the infrastructure already exists: Electrical outlets are abundant. Battery technology is also improving about 8% each year, North said, allowing bikes to easily upgrade once the chemistry comes along. Already, electric two-wheelers get the equivalent of about 300 to 500 miles per gallon. As technologies improve, they’ll be able to generate even more energy with less weight and cost.
The Price?
Most currently available production electric two-wheelers cost less than $10,000. Vectrix was the first company to manufacture a production electric two-wheeler. Since introducing its $11,000, 62-mph Maxi Scooter in August 2007. This spring Vectrix will roll out a third scooter model, the $5,195, 30-mph VX-2. Zero was the second manufacturer, after Vectrix, to make a production electric two-wheeler. Founded by Saiki, a former NASA engineer, and funded in part by former Sun Microsystems executive Gene Banman, who now serves as Zero’s CEO, the company has sold 200 of its $7,500 Zero X models — an off-road electric motorcycle with a 50-mph maximum speed and 40-mile range off a single charge. Brammo’s Enertia claims a top speed of 50 mph, a 35- to 45-mile range on a single charge of its lithium-ion battery pack and a $8,995-to-$14,995 price tag. The cheapest version reflects a battery-lease program that reduces the bike’s cost, bringing it more in line with similar, gas-propelled products.
I just came across a new website called routeRANK, which ranks a route based on transit mode (car, rail, or air), price, time and carbon dioxide emissions. It’s currently only offered in Europe (which makes good sense because train transit is a viable option to compete with short flights).
Here’s a routeRANK screenshot of Madrid to Barcelona in Spain sorted by price (the default)–click on the picture to get a higher resolution:
Here’s the same route, sorted by CO2 emissions–click on the picture to get a higher resolution:
Notice that train travel is by far the most sustainable form of travel according to routeRANK, by nearly an order of magnitude over comparable flight options. I wonder how many people would be persuaded to take a slower, but more environmentally friendly route based on this information? Another pertinent question is, how accurate are these CO2 emission estimates? According to the routeRANK website:
CO2 emission calculations are based on a model developed by the IFEU Heidelberg. They are further refined using information from the European Commission, non-profit organizations, transport providers and universities across Europe.
The calculations account for emissions generated by transport vehicles (e.g., gasoline and diesel from the vehicle’s tank), emissions generated by the extraction and conversion of energy (e.g., crude oil, coal, uranium from power plants or refineries), and emissions generated by energy distribution (e.g., tank trucks, power grid, oil tankers). They do not include emissions generated by construction, maintenance and disposal of transport vehicles (e.g., cars, planes, trains) or infrastructure (e.g., roads, airports, railway lines).
Car emissions are based on those of a mid-sized, gasoline-powered passenger car (EURO 4) with an average of 1.5 passengers. Plane emissions account for differences in capacity utilization where data is available. Similarly, train emissions account for differences in capacity utilization and consider national differences in electricity mix.
Users can customize their car by choosing the fuel type (petrol or diesel) and car type (small, medium, large, or custom car). The custom car option lets the user enter the exact fuel consumption of the car, which will then be used in the calculations.
Want to know even more?
We can provide you with a detailed sheet of CO2 calculations for the different forms of transport. Please contact us to receive this information or with any further questions you may have.
On a slightly unrelated note, when I was interning with John Krumm at Microsoft Research back in 2007, we brainstormed a research project that would rank driving routes based not just on time (i.e., the most efficient) but also on safety (e.g., reroute around dangerous intersections). In the future, I imagine such information being readily-at-hand to help us make more informed travel decisions.
KQED recently did a story on a pilot program in San Francisco where high school students are given Nokia cell phones to test a mobile carbon tracking system.
The way the San Francisco pilot program works is like this: students get a mobile phone equipped with a GPS maps application. They fill out a profile with the make and model of the cars they use. The cell phone monitors movement, so it picks up when that student is making a car trip. The server factors in the time of day, the weather and humidity, and the type of car the student is riding in – and then calculates the amount of carbon output that trip represents.
The program currently doesn’t differentiate between cars and other forms of transportation – bikes, ferries, trains, carpools, buses – so students may need to note when those trips were not regular car trips. The final number is their carbon rating.
When the program expands to three other San Francisco schools at the end of March 2009, a competition will be formed between the high schools to see which group of 25 students can cut back the most on their car trips and carbon output.
That will help answer the question of how much pollution people can save just by altering transportation behavior. And hopefully, the participants here are young enough that those transportation choices might continue after the program has ended. Once they get used to walking or biking, for instance, maybe they’ll make that a regular form of transportation.
Note that this is similar to the UbiGreen Mobile Transportation display; however, in that project we did not use GPS instead opting for a sensing platform that was capable of inferring walking, running, and bicycling in addition to driving in a vehicle.
Most people are unaware of how their daily activities affect the environment. Previous studies have shown that feedback technology is one of the most effective strategies in reducing electricity usage in the home. In this position paper, we expand the notion of feedback systems to a broad range of human behaviors that have an impact on the environment. In particular, we enumerate five areas of consumption: electricity, water, personal transportation, product purchases, and garbage disposal. For each, we outline their effect on the environment and review and propose methods for automatically sensing them to enable new types of feedback systems.
I had two primary goals in mind while writing this paper:
to inform the reader, particularly HCI practicioners and researchers, about the ways in which environmentally impactful human behaviors can be sensed
to inspire thinking about ways in which these new types of sensor data may be aggregated, analyzed, and fed back to the individual in order to increase awareness about environmentally impactful activities and motivate sustainable behaviors.
Needless to say, a paper such as this begs the question, even if we can automatically sense human activities that impact the environment, should we? Whenever we talk about sensing and automatic detection, Orwellian fears come to mind. These fears are certainly justified. My hope would be that human behavior data need not go beyond the user’s own device. This does not all together eliminate the problem (e.g., the device could be compromised) but certainly mitigates it. A better question is, perhaps: is sensing/feedback technology an effective strategy in reducing consumption? Prior studies in energy feedback technology have demonstrated that providing information about energy use to residents does often reduce consumption. Will this translate to other domains? What are the most effective ways in providing feedback? Does the feedback have to be persuasive or can it simply be informational? For other questions like these, see the paper.
V2Green is a local Seattle greentech company started, in part, by UW grad Seth Bridges (left in photo above). V2Green is based on a rather clever idea, it provides electrical grids the opportunity to store and draw power from plug-in electrical vehicles.
V2Green technology enables the flow of energy between electric vehicles and the grid to be adaptively managed, balancing real-time grid conditions with the need to charge individual vehicles.
The benefits are clear. Smart charging allows the existing grid to support electric transportation. Utilities, eager to increase their use of renewable energy and encourage off-peak charging, are expected to offer vehicle owners economic rebates. The resulting lower cost of ‘electric fuel’ will drive plug-in vehicle sales and encourage auto makers to further their investments in clean-energy transportation.
Ultimately, V2Green solutions support a reduction in fossil fuel consumption—decreasing the greenhouse gas emissions that cause climate change and diminishing the nation’s dependence on foreign oil.
These notes are from a talk given by Seth Bridges to an embedded systems class at the University of Washington on February 23rd, 2009.
Opening Slides
Is an electric vehicle any different than any other appliance that you would have in your house? On the face of it, maybe not. However, you don’t use the vehicle while it is plugged in–if you are, you’re probably doing something wrong. Air conditioners, pool pumps, refrigerators–even if you turn them all on–does the system fall over? No, it doesn’t. So, what happens when we plug in one more thing.
The average vehicle gets 4 miles / kWh. For the year, it’s about 3,000 kWh / yr (for 12,000 miles / yr). The average household is 10,000 kWh / year. So, is this a problem? Well, maybe no. Especially if you are charging at night; the problem is, however, if you plug in during the day or even at 4pm when you get back from work.
CA gives people $250 a year for installing a system that allows the utility to remotely control A/C. The rebate program is driven by Southern California Edison (link). How much would people expect to receive for taking part in GridPoint. Shows iPhone application. You need a way to opt-out, most certainly.
Grid 101: Reducing Peak Demand
One of the big things you have to think about is peak demand. If you plug in your car, you draw 10-15 amps. GM and Ford are saying that plug-in cars will have substantial charging requirements, something that makes them look like an air conditioner. The cost to run a coal plan, 2 cents per kw/hr. Natural gas 8-10 cents kw/hr. At the maximum peak demand, it can cost upwards of $10,000 per kw/hr. Last summer, CA hit 50 GW on August 11th (or thereabouts) and their capacity is 56GW. They were within 10% of maximum. This is why utilities are scared of thousands, or even millions, of plug-in cars. They can’t build that infrastructure fast enough. Our system allows utilities to control load on the electric grid by adjusting charge rate on PHEV vehicles. Eventually we will send power back into the grid from the vehicle battery system in order to minimize excessive loads on the power grid.
Grid 101: Real-time balance
Generally, there is no storage in the grid–that is why every moment they are balancing supply and demand. It’s really expensive to take a gas turbine to move it off its most efficient point; every time you take it down or modify it, you are wasting natural gas. If we can somehow save energy and match load via plug-in vehicle batteries, it would be a tremendous benefit. There is a lot of inefficiencies in the system to just provide real-time balance–there are systems on standby, long-term standby, etc.
Grid 101: Renewables Integration
Due to emerging state/federal regulations, 15-20% of all new generation will have to be renewable. This sounds pretty good; everyone likes wind and solar. However, these resources are variable. A 1MW wind turbine won’t put out 1MW all day, you can predict these things in aggregate but it’s challenging. If you have a cloud that moves over a 1MW solar farm and you lose around 1MW off the grid, that’s a big problem. If you could, for example, take your cars or water heaters to schedule load and charge only when wind and solar are pushing out more electricity, that would be most efficient. Is this feasible?
Opportunity: Manage Your Load
The Tesla requires 80 amps at 240. You need special box in your house to charge it. The neighborhoods that have Priuses will likely upgrade to Teslas (or cars like that). Power draw of everyday appliances: water heater is 4kW, TV is 150 watts, pool pump is 1-2kW. As long as you can drive your car, you’re not going to care when you drive it. The Chevy Volt is talking about 16volt battery pack, which they will probably use only 10 so they get long lifetime. It costs a lot of money for a charge cycle. Lithium Ion is about $2 per Wh–for the car though, people already have an incentive to buy batteries.
The technology that is in more sophisticated drive trains can absolutely drive 20kW right out the door. GM and Ford though don’t want this to mess up warranties–the 100,000 mile warranties.
Challenge: Manage Your Load
It takes lots of communication. Today, we have 130 cars. We get to most of them by using Verizon and AT&t, cellular modem contracts. It costs a lot of money to send data on that network, this won’t scale. Some of our customers have their own networks, like 100-200 bytes a day to the house. What does it cost to push a byte out, that has been an increasing point of pain. How do you send less bytes? How do you send bytes more cheaply? Unlike a water heater, the car gets up and goes away. At the end of the day, between the hours of 10-6, I will have X amounts of MW. This is about machine learning and predictive analytics. The better you are at this, the more money you can make.
V2Green Connectivity Module
Shows embedded system picture. Now on 3rd version of hardware. This piece of hardware communicates with cars (Priuses, etc.). It does the normalization. It might be serial port or CANN bus. Has GPS and data logging. Provides high grade energy monitoring. AC power supply, it can pull from vehicle or pull from wall. Uses everyday components. Atmel and 64 Mbit external flash. Lots of data logging. Up to 2/samples per second. Flexible communications daughtercard to support wifi, internal cellular, zigbee.
Open Source: Key Software Building Blocks
They use open source software but don’t release their software so they choose carefully.
FreeRTOS: simple, small preemtive scheduler (highly recommends it)
IwIP: a lightweight TCP/IP stack
axTLS: memory efficient TLS1 implementation
sustain is a blog dedicated to the environment, human behavior, technology and the relationship between all three. subscribe to the rss or atom feed.
Author
jon froehlich is a phd candidate in
computer science at the university of washington.
his research focuses on building and studying technology that promotes healthier lifestyles and proenvironmental behaviors.
