Clifford Winston: Paving the Way for Driverless Cars (Big Data)

     
Clifford Winston,  The Wall Street Journal, July 17, 2012

Instead of focusing on an enormously expensive high-speed rail system, government should promote modern highway design for cars of the future.


California's proposed bullet train between Los Angeles and San Francisco—which Gov. Jerry Brown is likely to sign off on soon—has been characterized by the Obama administration and its other supporters as an effective way to reduce highway congestion. These costs amount to more than $100 billion annually in wasted time and higher fuel expenses. 

In fact, a much better technological solution is on the horizon, if we pave the way by getting rid of obsolete highway design. It is already possible to imagine a world in which you could predict exactly how long it would take to drive in your car from one point to another. No worries about rush hour, vacation congestion, bad drivers, speed traps and accidents. You could also text while you drive with no safety implications. 

All this may be possible thanks to a "driverless" car that does a human driver's normal job and much more. The car is operated by a computer that obtains information 10 times per second from short-range transmitters on surrounding road conditions, including where other cars are and what they are doing. 

That's exponentially faster than the human mind can process the same information. By gathering and reacting immediately to real-time information, the technology can drastically reduce highway fatalities by preventing collisions. It also can significantly reduce delays by creating a smoother traffic flow and rerouting drivers who have programmed their destinations. 

Google's version is being piloted in Nevada, and it could prove that faster, reliable and safer road travel is within reach. But one stumbling block would remain: the government-run roads this innovation must use. Auto makers have made one technological improvement after another since the car was introduced to consumers more than a century ago. Unfortunately, the paved road systems on which cars travel have not advanced much in comparison. Without reimagining the way we design and maintain highways, the driverless car will achieve little of its potential. 

Despite the frustratingly frequent lane closures for repairs, about one-third of the nation's highways are still in poor or mediocre condition. Driving on damaged roads is hard on vehicles and is estimated to cost motorists billions of dollars annually. Those potholes could also defeat the purpose of the driverless car because it would be unable to avoid them, or succeed in doing so only by significantly disturbing the traffic flow.

Most highways in major metropolitan areas operate under congested conditions during much of the day. Yet highways are designed around standards based on higher free-flow travel speeds that call for wider but fewer lanes. Driverless cars don't need the same wide lanes, which would allow highway authorities to reconfigure roads to allow travel speeds to be raised during peak travel periods. All that is needed would be illuminated lane dividers that can increase the number of lanes available. Driverless cars could take advantage of the extra lane capacity to reduce congestion and delays. 

Another design flaw is that highways have been built in terms of width and thickness to accommodate both cars and trucks. The smaller volume of trucks should be handled with one or two wide lanes with a road surface about a foot thick, to withstand trucks' weight and axle pressure. But the much larger volume of cars—which apply much less axle pressure that damages pavement—need more and narrower lanes that are only a few inches thick. 

Building highways that separate cars and trucks by directing them to lanes with the appropriate thickness would save taxpayers a bundle. It would also favor the technology of driverless cars because they would not have to distinguish between cars and trucks and to adjust speeds and positions accordingly.

Traffic management also suffers from obsolete technology that could hinder implementing the driverless car. On local streets, signal timing contributes to hundreds of millions of vehicle hours of annual delay because it is based on out-of-date historical data that inaccurately measure relative traffic volumes at intersections. Without signals based on real-time traffic flows, driverless vehicles may not be able to accurately align their speeds with them.

The future also holds the promise of new communications technologies that could let road authorities use electronic tolls to charge motorists for their contribution to congestion, based on actual traffic conditions, and thus encourage them to travel during off-peak periods, use alternate routes, or switch to public transit. Driverless cars would significantly help motorists respond to congestion tolls because their technology can balance the cost of a toll with its travel time savings to optimize motorists' route choices. 

The driverless car represents one of the most amazing breakthroughs in safety and quality of life in recent history. Instead of focusing on enormously expensive high-speed rail as our transportation future, the government would do well to stop hindering driverless cars by its obsolete thinking about our nation's roads. 

One promising approach that would not require taxpayer funds would be to turn to innovative private highway companies, which have leased the Indiana toll road, Chicago Skyway and Dulles Greenway. By working closely with auto makers, they could significantly shorten the time that motorists must wait before they fully realize the benefits of driverless technology. 

Mr. Winston is a senior fellow in economic studies at the Brookings Institution and author of "Last Exit: Privatization and Deregulation of the U.S. Transportation System" (Brookings Press, 2010).

 

A version of this article appeared July 18, 2012, on page A15 in the U.S. edition of The Wall Street Journal, with the headline: Paving the Way for Driverless Cars.

Better Medicine, Brought to You by Big Data

Derrick Harris, Gigaom, July, 15, 2012

 

 Slowly but surely, health care is becoming a killer app for big data. Whether it’s Hadoop, machine learning, natural-language processing or some other technique, folks in the worlds of medicine and hospital administration understand that new types of data analysis are the key to helping them take their fields to the next level.

Here are some of the interesting use cases we’ve written about over the past year or so, and a few others I’ve just come across recently. If you have a cool one — or a suggestion for a new use of big data within the healthcare space — share it in the comments:

Genomics. This is the epitomic case for big data and health care. Genome sequencing is getting cheaper by the day and produces mountains of data. Companies such as DNAnexus, Bina Technologies, Appistry and NextBio want to make analyzing that data to discover cures for diseases faster, easier and cheaper than ever using lots cutting-edge algorithms and lots of cloud computing cores. Dell is providing computing power for two research centers to try and treat a particular form of pediatric cancer based on each child’s specific genetic profile.

BI for doctors. Doctors and staff at Seattle Children’s Hospital are using Tableau to analyze and visualize terabytes of data dispersed across the institution’s servers and databases. Not only does visualizing the data help reduce medical errors and help the hospital plan trials but, as of this time last year, its focus on data had saved the hospital $3 million on supply chain costs.
 

 
Semantic search. Imagine you’re a doctor trying to learn about a new patient or figure out who among your patients might benenfit from a new technique. But patient records have been scattered throughout departments, vary in format and, perhaps worst of all, all use the ontologies of the department that created the record. A startup called Apixio is trying to fix this by centralizing records in the cloud and applying semantic analysis to uncover everything doctors need, regardless who wrote it.
 

 
Hadoop for everything. Cloudera is partnering with the Mount Sinai School of Medicine to help it develop new methods and systems for analyzing biological data. But that’s just the latest of Cloudera medical efforts, which also include working with the Food and Drug Administration to detect unsuspected adverse side effects from multi-drug combinations, and Emory University on helping pathologists more accurately analyze medical images. One of Cloudera’s customers, Explorys, built a business around aggregating and analyzing medical records, and Intel and NextBio are teaming to tune Hadoop for processing genomic datasets.

 
Watson. IBM has dozens of irons in the healthcare fire, but its coolest might well be a partnership with WellPoint to put the Jeopardy! champion question-answering system in doctors’ offices. Watson could help doctors answer questions posed in natural language by analyzing them against mountains of medical research data that no individual doctor could possibly read and digest.

 

Getting ahead of disease. It’s always good if you figure out how to diagnose diseases early without expensive tests, and that’s just what Seton Healthcare was able to do thanks to its big data efforts. Trying to find better ways to detect congestive heart failure early in order to save the exorbitant costs of treatment as the disease progresses, a team found that a distended jugular vein — something that can be spotted during any routine physical exam — is a particularly high risk factor.

 
Data scientist in residence. Here’s a new title for a healthcare organization — chief data scientist. Yet, that’s exactly the position Alliance Health Networks just added in May. The company, which provides social networks focused on specific medical conditions, acquired medical research database Medify and decided it needed someone to lead the effort of analyzing all that data and providing valuable feedback to network users.

 

 
Crowdsourced science. In a field where controlled experiments can be expensive and sometimes ineffective, it’s turning out there might be no substitute like the real-world data. Probably the most widely known company in this space is PatientsLikeMe, a social network designed to let individuals share their medical conditions so they can learn from others like themselves what treatments might work best in their particular circumstances. As a side effect, the company is able to conduct observational trials based on data users willingly volunteer.

 Feature image courtesy of Shutterstock user lenetstan; Tableau graph courtesy of Perceptual Edge; exam image courtesy of Shutterstock user Blaj Gabriel

The Way the Digital Cookie Crumbles

If regulators and lawyers limit the use of data, advertising online will become less efficient. 

 

L. Gordon Crovitz, The Wall Street Journal, July 15, 2012.

For a measure of how technology is changing human expectations, consider the "cookies" on your computers. These invisible text files are how websites track activity, delivering to marketers detailed information about individual behavior and preferences. In exchange for data, we get highly personalized online services. 

This use of cookies fuels the economics of the Web, but it has also caused anxiety as people have had to reconsider analog-era expectations of privacy to embrace digital-era benefits of sharing data. A Wall Street Journal report last week caused some consternation when it revealed how the travel website Orbitz uses data to give different offers to people who use Apple computers and those using Windows-based machines. 

Data analysts at Orbitz detected patterns showing that Apple users spend up to 30% more a night on hotels and are likelier to book four- or five-star lodgings than PC users. Apple users buy more expensive computers, and the average household income for adult owners of Mac computers is almost $100,000, compared with about $75,000 for PC owners, according to Forrester. 

When Orbitz used these data to feature higher-priced hotels more prominently in Apple users' search results, privacy lobbyists claimed outrage. But even in the analog era, readers of this newspaper saw advertisements for different products and services than readers of less high-end papers. 

It wouldn't be surprising if data showed less price sensitivity among Apple users, so they could be offered higher prices, but at least for now Orbitz shows the same prices for the same rooms regardless of how users access the site. Price-customization software is being used by many retailers so that online buyers who click directly to pay for products do not get special offers shown to less-persuaded shoppers.

These uses of personal data can seem a bit creepy, but the evidence also shows how quickly consumers have gotten used to being tracked. When given the choice, few consumers opt out of cookies. People accept the benefits of more relevant ads and more personalized websites in exchange for letting marketers track their interests. 

There is an enormous industry in predictive analytics and "big data." Consumers are loyal to Amazon in part because of its recommendation tools—if you liked that book, you may like this one—which mine user data to determine relevancy. Facebook says it will deliver targeted ads based on what other websites and apps users access. Apple tells users it will target ads to them based on apps they download. Google delivers advertising based on how people use its various services, including what they write in messages sent via Gmail.

Left alone, people would continue to make their own evolving judgments about how much data to share. Instead, regulators issue edicts. The Federal Trade Commission has extracted 20-year consent decrees from Google, Facebook, Twitter and Myspace, giving regulators broad review over their privacy and data practices. This would be fine if the purpose were to ensure that companies comply with disclosures about how they use data, but the FTC wants to define privacy standards. 

One result of FTC meddling is that plaintiff lawyers have open invitations to file nuisance suits on behalf of supposed privacy victims. A federal judge is considering a $20 million settlement offer by Facebook, which has agreed to make its disclosures clearer that when users click "Like" to promote a product on Facebook, their names and photos can be used. 

The $20 million would be divided equally between plaintiff lawyers and privacy interest groups such as the Electronic Frontier Foundation. Nothing would go to the allegedly harmed 900 million Facebook users.

"The plaintiff's lawyers get rich, class members get little and nonprofit groups often reap millions by urging judges to approve the deal regardless of its merits," Wired magazine reported last week on the Facebook settlement. This case "provides a glimpse into the dark side of large class-action settlements."

If regulators and lawyers push too hard to limit the use of cookie data, advertising online will become less efficient. This in turn will reduce the amount of free, advertising-supported services enjoyed by consumers, such as social media, entertainment and email. 

Consumers seem to understand there's no such thing as a free lunch, even online: If they are not paying for a product, then for better or worse, they are the product. Each consumer should be able to decide how to make this trade-off between sharing data and getting advertising-supported services.

The privacy debate shows how naive Silicon Valley firms were to sign 20-year agreements granting Washington regulators broad authority over how they operate. Digital entrepreneurs should be allowed to innovate freely, with consumers also free to choose their individual trade-off between how their data are used and the benefits they get in return. Overregulation is the way the digital cookie crumbles. 

A version of this article appeared July 16, 2012, on page A11 in the U.S. edition of The Wall Street Journal, with the headline: The Way the Digital Cookie Crumbles.

The Age of Big Data

Brad Peters, Forbes, July 12, 2012

Big Data, the application of a new breed of analytical tools to the vast caches of data being produced by computers and other forms of technology, is on the brink of becoming a household word – thanks to new books, conferences and articles that are slated to appear over the next few months.

But as impressive as these Big Data stories will be – and many of them, like mapping the insides of F5 tornados, tracking every patient heartbeat over a lifetime and predicting the behavior of consumers from millions of store purchases – the real story is unlikely to be told.

That is, Big Data has the potential to utterly transform the relationship that individuals have with institutions, customers with companies, patients with the healthcare system, students with universities, and voters with government. And that means once it has fully penetrated society and industry, the Big Data revolution may very well prove a turning point in our economic – and ultimately, cultural – history as great as the electronics revolution. . . perhaps even as great as the first and second Industrial Revolutions.

Why? Because once the relationship of individuals to institutions transforms, the benefits to the individual consumer, citizen, patient and student will be profound.

To explain why this will be so, let’s look more closely at the impact of Big Data on the commercial world, where this revolution is likely to occur first.

The story of business over the last two centuries can be seen as the evolution of the relationship between sellers and buyers. The first Industrial revolution, at the end of the 18th century, brought mass production of standardized goods – and thus greater volume and lower cost – to the masses, supplanted the high-cost, customized goods that had characterized commerce for millennia. The second Industrial Revolution, beginning in the middle of the 19th century, brought that mass production to the creation of interchangeable components – meaning that those parts could be swapped out and pieced together into finished goods almost everywhere. . . meaning even more availability and lower prices to everyday consumers.

The digital/electronics revolution, by putting intelligence into the production process – even into the finished goods and services – made possible for the first time mass customization. That is, manufacturers and providers could create products and servers in high volume and reduced cost, and those creations could then use their on-board intelligence to adapt to the buyer’s unique needs. Think laptop computers, cellphones, the Internet, user customizable automobiles, etc. The extraordinary richness of modern life – especially as it has reached out to include three billion of the world’s people – can be largely credited to the mass customization revolution.

But now, Big Data – and the billions of high-speed, low-cost processors, computers, sensors and networks that make it possible – promises to take this relationship to the next level: mass personalization. That is, it uses all of this computing power, and the mountains of data now being produced by each person each day, and uses it to continuously and uniquely modify products and services to every customer.

And that’s only the beginning: because in this pursuit of the perfect 1-to-1 relationship with each customer, the very nature of that relationship begins to change – not least because it becomes part of the duty of the supplier to keep that relationship continuously optimal. And to do that, the manufacturer or service supplier can no longer (as with mass customization) ask for a one-way flow of information from the customer. Now for the first time it must send valuable information the other way. They have to share with customers what they know in order to get further information on what those customers want – the information they need to make their Big Data analytics even more accurate.

A pioneering example of this is Facebook. The only way that it can make money and continue to grow (and reward those frustrated new shareholders) is to personalize their offering so completely to each of its near-billion users that those users trust it so much that they give up ever more layers of personal information. . . which can then be analyzed and sold to industry. For example, again thanks to Big Data, every Web ad you will soon see on Facebook and across the Web will have been bid upon in real time by advertisers who will pay based upon your perceived value as a potential customer.

As we’ve seen with some of Facebook’s failed attempts in the past to do this, it is not an easy task. But the reality is that in our technology- and media-saturated world, they have no alternative – nor do most companies, whether they know it now or not. The new world of mass personalization requires relationships built upon mutual trust. And more than at any time in history, companies – and soon, big government, big education, big everything – are going to have to work hard to earn that trust from each and every one of us. The world now has to come to you and me as consumers – not the other way around.

That’s why Big Data is so important. It changes the social contract. And the more influential it becomes, the more its tools will be in demand. Ready or not, we are entering the Age of Big Data.

The Power of Information

Greg Satell, Innovation Excellence, July 10, 2012

What’s the big deal about the information economy anyway?  Surely, information has played a role in commerce since ancient times.  What’s changed?

One reason for the confusion is that we’re not used to making the distinction between information and knowledge.  Go to Istanbul’s ancient Grand Bazaar and you’ll instantly grasp that the traders know a lot, but not much that they can easily share even if they want to. That, it turns out, makes all the difference.

The emergence of information as a storable, fungible entity is transforming our economy and our society in ways that we scarcely realize.  It’s making us richer, smarter and even healthier.  What’s more, its impact is accelerating, so we’ll see a lot more change in the coming decades than we have in the past.  In fact, we’re just getting started.

A Theory of Information

If you had to put a date on it, the digital age truly began in 1948.  It was in that year that two seismic events happened, both at Bell Labs.  The first and the more famous was the invention of the transistor, which forms the basis for our present (albeit soon to be defunct) digital technology paradigm.

The lesser known, but in some ways more important event, was Claude Shannon’s groundbreaking paper A Mathematical Theory of Communication, which launched information theory out of thin air, seemingly with no precursor.

Although not widely publicized or understood at the time, it has become central to all the digital technology we use today.

The basic idea was that information isn’t a function of content, but rather a lack of randomness, which can be broken down to a single unit – a choice between two alternatives.  Much like a coin toss which lacks information while in the air, but takes on a level of certainty when it lands, information arises when ambiguity disappears.

He called this unit, a “binary digit” or a bit and much like the pound, quart, meter or liter, has become such a basic unit of measurement that it’s hard to imagine our modern world without it.

Storage and Transfer

In the late 1940’s, Shannon’s colleague at Bell Labs, Richard Hamming became frustrated that computer errors continually ruined his work.  He built upon Shannon’s paper and created Hamming code, small bits of extra information that would allow computers to detect and correct errors.

Of course, that increased the amount of information that needed to be processed.  No problem, information theory also shows us how to compress information by eliminating redundancies.  Common technologies that we have come to use everyday, like JPEG and MP3 are, in fact compression techniques that have their roots in Shannon’s 1948 paper.

It is storage and transfer that make the information age so different from what we knew in the past.  In contrast to the knowledge that a bazaar trader possesses, computers can duplicate and transfer information an infinite number of times, with as little error as we choose.

Accelerating Returns

The ability to store and transfer information efficiently has an interesting side effect – we can improve at an exponential pace.  Returns to our efforts not only increase, they accelerate.  The most famous example is Moore’s Law, which says that processing speeds double about every 18 months.

 

Look at the chart above and notice the logarithmic scale.  From 2000 to 2010, the number of transistors on a single chip increased from 10 million to a billion.  That’s a hundred times more than the increase the previous decade and nearly a million times more than the decade before that.  In the next ten years, we can expect it to increase 100 -fold again.

What’s even more amazing and also of paramount importance, is that the principle isn’t exclusive to processing, but applies to every facet of information technology, from storage to bandwidth to power consumption, everywhere you look, efficiency continues to improve exponentially.

The Information Invasion

Here’s where it gets really interesting.  As information technology becomes more widely deployed, the information content of other products and services increases and they begin to follow the same exponential trends.  Take a look at genome sequencing:

As sequencing genomes became less of a pure biological science and more of an information science, the pace of advancement changed drastically enabling a whole new field of bioinformatics that will revolutionize medicine.

Similar trends are being played out across almost every industry you can think of.  In manufacturing, new technologies like 3D printing and (eventually) programmable matter are creating what The Economist calls a third industrial revolution.  Even when you go and buy a box of cereal at Wal-Mart, a good portion of the retail price is made up of informationally dense logistics.

Ray Kurzweil spoke volumes when he said that in the future “all technologies will essentially become information technologies, including energy.”

From Belief to Observation

As the informational content of products and services increases and returns accelerate, lots of good things happen.  Science fiction becomes engineered fact.  The unthinkable will become commonplace.  Incomes will rise while poverty falls.  Seemingly intractable problems will be solved and dire needs will be met.

Yet there is also quite a bit that is unsettling.  As technological cycles shorten, business models will have shorter life spans. The internalized experiences that we have come to regard as intuition will fail us more often.  Our ability to plan will diminish and the need to experiment (and fail) will increase.

And that’s what’s disconcerting.  This new information economy doesn’t run on beliefs or even, to a certain extent on ambition, but algorithms which, powered by ever more abundant processing power, test and accept or discard a dizzying multitude of possibilities, the results of which can be retrieved and recombined with other experiments.

The power of information means that we are no longer required to believe, only to imagine, test and observe.

Amazon economy: From warehouse to powerhouse

Barney Jopson, The Financial Times, July 8, 2012

Once just a book seller, the company provides vital logistics for many but it seems set to face sterner political scrutiny 

GoVacuum is a fast-growing ecommerce business whose annual sales have gone from under $2m seven years ago to $10m today, a fivefold expansion that surpasses the overall growth of US internet shopping. Masterminded by founder Bill Anand, who arrived from New Delhi in 1974, such growth underlines the importance of small business to the world’s biggest economy and commands kudos in an ecommerce sector where Amazon seems to hog the spoils.

But the success of GoVacuum and its 27 staff has not come in spite of the online behemoth from Seattle. It has come because of Amazon.

Most consumers still view Amazon as an online book retailer. Some are surprised to find it sells much more than the single product with which Jeff Bezos, its founder and chief executive, started in 1994. But the store front is just the tip of an iceberg.

Behind the scenes, Mr Bezos is turning Amazon into a back office infrastructure provider that sells access to a digital market place with millions of customers, to petabytes of server space and to state-of-the-art warehouse facilities serving myriad forms of commerce, including GoVacuum’s.

“They are investing in areas that touch so many people’s lives they’re becoming about as important as utilities,” says Sachin Anand, GoVacuum’s business development manager and Bill’s son. 

Amazon has long thrived by overturning the way people shop, but its shift into infrastructure is extending its power as a disruptive force to how business is structured. It is revolutionising the way entrepreneurs can create start ups, or revive staid companies, by letting them plug their ideas into pay-as-you-go systems that cost a fraction of the investment they would need to build such infrastructure alone. “They allowed us to achieve a scale that on our own we wouldn’t have been able to,” says Sachin Anand.

This has lifted Amazon’s economic influence beyond its tech peers Apple, Google and Facebook and taken it into the realm of network businesses such as stock exchanges, power grid operators, credit card processors and shipping lines.

But its emergence as a new and largely unregulated steward of such vital commercial plumbing creates tensions and dangers for clients and parts of the US economy. 

Individual companies must weigh the benefits of joining Amazon’s “ecosystem” against the costs of losing autonomy and the dual role Amazon plays as both a rival retailer and a service provider that is able to see their sales and inventory – critical details for any business.

At a macro level, hub businesses such as Amazon improve efficiency, which is good news for small players, says Marco Iansiti, a professor at Harvard Business School. “The bad news is your destiny is shared with a bunch of other people,” he says. “If Amazon has a catastrophe, if the website goes down, then a lot of people are affected. It is a source of systemic risk that we didn’t have before.”

Amazon is notoriously opaque and its shift into services is making a growing proportion of its business less conspicuous to the public eye, just as it becomes more critical than ever to the US consumer economy.

Unbundling the Corporation

In a series starting on Monday, the Financial Times is charting Amazon’s ability to reinvent industries, drawing on over 50 interviews with former employees, clients, software providers, rival retailers, academics, bankers and consultants.

What emerges is a voracious company whose no-frills technology and pragmatism have helped it build a commanding role in ecommerce, while also shaping its reaction to strident criticism over not collecting US sales tax, its first big political battle.

But Amazon has flaws: it remains heavily dependent on its founder, it has not kept up with the advance of digital music and movies, and it has yet to convince many investors that it will ever make a substantial profit. Its growing influence is also likely to face more political scrutiny. 

In his latest letter to shareholders in April, Mr Bezos said: “We are creating powerful self-service platforms that allow thousands of people to boldly experiment and accomplish things that would otherwise be impossible or impractical.”

Its services have already produced hybrid businesses where Amazon runs marketing, customer relationships, payments, computing, logistics and distribution, leaving executives to do nothing but find or make good products, both physical and digital.

In effect Amazon is “unbundling” the corporation – a concept coined by McKinsey management consultants in the late 1990s – by letting companies shed supposedly “core” processes that conventional wisdom says should be combined within a single entity. Doing so, however, ties their cost bases inescapably to Amazon.

A key attraction for Amazon, which made a meagre $631m net profit on revenue of $48bn last year, is that it is more profitable to earn service fees than to buy and sell products itself.

Beyond pure retail, one per cent of all consumer internet traffic touches Amazon-managed infrastructure, according to DeepField Networks, a computing group. Amazon’s Kindle Direct Publishing platform has enabled writers to produce ebooks.

Amazon Web Services, a cloud computing business, has become an engine behind digital content companies including Spotify, a digital music service, and Netflix, a video streaming service. Outages, however, have downed some client websites twice this month and last April. Spotify and Netflix also compete against Amazon’s Cloud Player and Instant Video services, but Amazon says information about cloud clients is not shared within its empire.

Still, unease among some large retailers about using Amazon services is evident. Target, a US discount chain, last year stopped running its website on Amazon’s systems and brought the job in-house. Marks and Spencer, a UK retailer, is doing the same. Toys R Us dropped Amazon as its webstore operator in 2006 after a judge ruled it had broken a contract by selling toys from other merchants on Amazon.com. 

As overlord of a mini-economy, Amazon must tread a fine line. “How do you create a healthy ecosystem in which you are not the only guy making money?” says Suresh Kotha, a professor at the University of Washington in Seattle.

“At some point people will switch if you become too powerful and start milking them,” he says. “If this happens … the ecosystem will slowly start to lose momentum.”

Mining clients’ data

GoVacuum last year sold its final store, in a small town in Virginia. The business that remains is partly fused with Amazon. Without it, says Sachin Anand, GoVacuum would have fewer sales, a narrower product range, a more unwieldy warehouse and less satisfied customers.

Between 40 and 50 per cent of GoVacuum’s revenue comes from Amazon’s market place, an online bazaar where it pays a 15 per cent commission on each sale to access US consumers – 85m unique users in May alone, according to ComScore research – who use Amazon as the Google of product search and visit its site at a rate of 32 a second. 

The deliveries are done by Fulfillment by Amazon, a distribution service, which is why the hoovers on the storehouse pallet are just passing through GoVacuum’s drab facility en route to an Amazon warehouse. They will be stored and shipped to customers for fees equal to roughly 10 per cent of each sale, depending on size and weight – a rate that is cheaper than any of the alternatives.

Amazon processes the payments of GoVacuum customers, keeps their email addresses to itself, and handles after-sales support. GoVacuum even uses the databases of Amazon Web Services to store information.

But Amazon’s commitment to its own front-end retail business is not waning. Sachin Anand has watched it appearing to treat GoVacuum as a laboratory where it can identify popular products then start to sell them itself at lower prices. “There are vague frustrations with the market place that, because they have visibility into what you are selling, they can quickly become your competitor,” says Mr Anand. “There’s no part of the agreement that says they can’t do that. When we started, Amazon wasn’t in our category at all.”

It should be no surprise. Amazon is famously hard-nosed and an unrivalled data miner, enjoying a privileged view of hordes of shoppers buying everything from detergent and shoes to pet food and patio furniture.

Diane Buzzeo, the founder of Ability Commerce, a software group that retailers use to manage their presence on Amazon, says: “They do look at your product mix and then go to the manufacturer and say: ‘Hey, why don’t we sell this direct?’ It’s amazing they haven’t been hit with some kind of non-competition lawsuit.”

One toy seller who did not want to be named says: “The work we and others have done has paved the way for Amazon to understand what sells and what doesn’t.” He tries to stay ahead by finding new designs and new manufacturers. “We have to protect ourselves against Amazon, which has perfect knowledge of everything.”

By contrast, eBay, a rival, uses the fact that it does not have its own retail business as a selling point.

In the first quarter, Amazon said its 2m-plus registered sellers – ranging from companies to dads emptying their garages – accounted for 39 per cent of unit sales by volume, up from 33 per cent a year ago.

Not all of them are bothered about competing against Amazon, which declined to make any of its executives available for comment. Jack Sheng, co-founder of eForCity, a seller of electronic accessories, says it does not unfairly monopolise customers and that, on its site, data on the most popular products and every seller’s line-up is publicly available. “Everyday I can tell you all our competitors are looking at our sales and entering our space,” he says. “Just suck it up.”

One investment banker says Amazon’s position is reminiscent of Goldman Sachs’ dual role as a broker and trader at the centre of capital markets. “People complain about conflicts of interest. But you still have to do business with them.”

Why services make sense

Like Goldman, Amazon has a ruthless reputation. This was sharpened in the eyes of Main street rivals last December when it gave a discount to shoppers who used its smartphone app to record a product’s price in a store then bought it from Amazon instead.

But its move into infrastructure has received little scrutiny, partly because unlike the cool products of Apple, Google and Facebook, Amazon focuses on dull-but-difficult tasks such as managing product listings and delivering boxes.

These are the foundation of its services. Shel Kaphan, who was Mr Bezos’s first Amazon employee and left in 1999, explains the shift in its business by noting that “platforms” are a favourite concept of software people and also hark back to Mr Bezos’s time at DE Shaw, a quantitative hedge fund in New York.

“Since Jeff came from Wall Street and worked at a company that was at the middle of markets the concept of a market place was always floating around. It’s the idea that being in the middle of everything helps you scoop up pennies and nickels,” he says.

With Amazon’s operating margin a mere 1.5 per cent in the past quarter – compared to Walmart’s 5.7 per cent – the shift into services could also solve what hedge fund manager David Einhorn has alluded to as the Amazon “riddle”: can it make serious profits?

In addition to low prices and free shipping, huge investments in warehouses and data centres have kept a lid on profitability. Its services enable it to make money from them.

Colin Sebastian, analyst at Robert W Baird, says overall Amazon probably breaks even selling products from its own inventory, but makes an operating margin of around 10 per cent on third party sales.

Mr Bezos’s talk of creating services to help entrepreneurs also has a whiff of venture capitalism, says Joel Bines, a consultant at AlixPartners. “But he’s not backing 10 companies, he’s backing thousands. And when one of them becomes a massive success, he continues to take his toll.”

Dependence discomforts

One such business, RJF Books and More, gets 100 per cent of its sales from Amazon and consists simply of its systems plus Bob Frank and his wife. From homes in Minneapolis and Phoenix, the couple drive around America buying interesting merchandise – from Tudor history books to bath soap – which they load into boxes in their car and send to an Amazon warehouse when full.

Mr Frank started the business after he was laid off as a software salesman and raised $1,000 by selling his own books on Amazon. Now his annual revenue is in six figures.

It would have been impossible without Amazon taking care of the selling, shipping, customer service, payments and complaints. “Sometimes I wish there was some other competition, because competition is always good … But no one else can do it as well as they can,” Mr Frank says.

Amazon says it focuses on helping sellers compete with other retailers, itself included, but the harsh truth is that most on its market place are expendable because someone can always fill their shoes. The sellers sign up freely and no one is accusing Amazon of breaking any laws.

The law, however, has not caught up with the evolution of technology platforms and competition lawyers are eternally vigilant. 

“In any of these models there’s a tension,” says Prof Iansiti at Harvard. “How you behave makes a big difference. Are you adding value to the ecosystem? Are you abusing your market power?”

Platforms tend to be winner-takes-all games where one contender eventually reaches a critical mass and everyone then gravitates to it. Amazon’s scale means entrepreneurs are finding it increasingly attractive to join and harder to leave, but signs of wariness among some businesses are growing.

GoVacuum, for example, is sprucing up its own website and building its presence on other market places run by eBay, Sears and Best Buy.

“There is the concern of the dependency issue,” says Mr Anand. “It would be easy to say ‘Hey, we are only going to focus on Amazon’, but there has to be some diversification.”

Mr Bezos has created a transaction engine that is reinventing the way entrepreneurs can do business. If his ambition is to colonise the entire infrastructure of consumption, it is not likely to be Amazon’s technological skills that set the limits, but the world’s willingness to accept him.

The "Algorithmic Revolution" in healthcare.

How Connected Health, Public-Private Cooperation, And Big Data Can Revolutionize Health Care

Guest Post by Dr. Jody Ranck,  Forbes, July 6, 2012

 

As we contemplate the fallout from the Supreme Court’s decision in NFIB v. Sebelius on the Obama Administration’s Affordable Care Act, it is useful to think beyond the din of the punditry on what the decision means in an electoral cycle and consider seriously the looming crisis in healthcare.

Our health system—or better, anti-system—consumes nearly $2 trillion annually and does not deliver the value that it should.  Those who think we have the best system in the world come up against the cold hard facts (if they matter in political debate anymore) of health outcomes, which indicate that we’re ranked 37^th in the world and trending downward, not improving.  Despite one of the most robust ecosystems for innovation in biotechnology and information technology that the world has ever seen, converting this into tangible health outcomes is an “innovation space” that we’ve yet to get a handle on.

There is hope and promising signs that these tools can be mobilized to make a dent in some of the major challenges we face such as obesity, diabetes, environmental health, and a host of other areas if we can bring politics, technology, and both the public and private sectors into the right relationship in the coming years.  

Technology can’t do it alone.  Nor will antiquated political hubris of “death panels” and market fundamentalism fix health care.  In my new book, “Connected Health: How Mobiles, Cloud and Big Data Will Reinvent Healthcare”, I explore how this might happen if we can think more strategically and engage in more grassroots efforts to engage with the “Algorithmic Revolution” in healthcare.

Mobile Health (mHealth)

Today we find mainstream publications almost daily reporting on new mobile health, or mHealth, apps to help you manage your fitness, diet or health condition.   There are over 10,000 health apps in the iTunes store, making them one of the fastest-growing categories, and it is estimated that 2012 will reach at least 247 million people who have downloaded at least one app, up from 124 million in 2011, according to the analyst firm Research2Guidance.  

The firm estimates that the world market for mHealth apps will reach over $1.2 billion this year.  These apps run the gamut from diabetes, exercise and fitness apps (Runkeeper, Fitbit), dieting (Lose It; the wireless weight scales designed by Withings), heart rate monitors (Azumio), sleep trackers (Zeo), mood trackers (Mood Tracker) and a host of peripherals that attach to your phone to provide everything from diagnostics for diseases to EKGs, eye exams (EyeNetra), and even microscopes.

Some mHealth apps use game dynamics to motivate collective action as in the case of the University of California-based CITRIS’ Pwning Asthma Triggers app that helps residents of high-risk neighborhoods identify the sources of industrial pollution that trigger asthma.  One of the most interesting startups coming out of MIT is Ginger.io, a platform that enables you to combine patient data with passive data collected on one’s cellphone.  Passive data provides insights on how individuals are using the device and can help one better understand social and psychological factors that can determine outcomes.   There are many who are looking to mobiles to help improve one of the sticky problems of healthcare: behavioral change that is often required to manage multiple chronic diseases.  Successful mHealth apps will have to demonstrate the ability to move the dial on behavioral change to warrant investment in coming years.

The Internet of Things, Health 2.0 and Mobiles as Health Hubs

Beyond mobile devices, there is the rapidly growing use of sensors that are connected to the Internet.  The Internet of Things, as it is commonly known, can also play an important role in future health systems.  From real-time air pollution sensors (the focus of a recent EPA Challenge, MyAir, MyHealth) and radiation sensors (Cosm (formerly Pachube)) to the growing market for wearables (Misfit Wearables, Zephyr Technology), we’re increasingly entering the early years of the Health Internet of Things.  That is, sensors connected to the Internet that can transmit data related to health.   When you combine this with the use of social networks for health, including platforms such as PatientsLikeMe.com, CureTogether, and general use of Twitter and Facebook for exchanging health information or organizing campaigns for health, the amount of health data and networks related to health and healthcare is exploding.

There is also the increasing number of individuals who use a variety of devices and social media platforms to track data about their health and fitness, experiment, and exchange insights.  Widely known as the Quantified Self, this movement is rapidly moving beyond early adopters in Silicon Valley and beginning to have an influence in even how consumer device manufacturers think about future health and fitness devices.  Qualcomm Life, perhaps anticipating the mainstreaming of aspects of the Quantified Self movement, has focused on creating the technology to integrate the range of devices in a simple, easy-to-use hub, the 2net Hub, that can make it easier for consumers to connect devices and data in their homes.

A Tipping Point: Democratizing Health Through Participatory Medicine

The exciting part of this ecosystem is not just the range of devices and impressive range of indicia we can monitor, measure, and analyze with the devices in our pockets, many of which did not exist with this level of functionality even 4-5 years ago.  Increasingly the Health 2.0 movement, or those interested in the use of social media and connected health to radically change the traditional top-down orientation of healthcare, view these technologies as a way to democratize health and medicine.  Calls for participatory medicine, or more patient-physician and even health system collaboration, are frequent, and there is even a Society for Participatory Medicine to debate and share information on the movement.  “ePatients” are challenging the traditional medical encounter armed with data they’ve tracked on a mobile or iPhone app and with the latest research or information gleaned from platforms that provide access to clinical trials and data.  Sometimes referred to as “cyberchondriacs”, the message is becoming clearer that we are reaching a tipping point where the practice of medicine and public health are going to change for good.  Information and data flows are no longer a one-way street.

Government and the Need for Public-Private Cooperation

In contrast to many in the field of health technology, I do not necessarily see a single trajectory for where this is heading.  There are many major technological, social, and political roadblocks and obstacle courses to navigate in the coming years.  First, the most difficult challenge is the political rhetoric in the United States concerning healthcare.  The recent Supreme Court decision on the Affordable Care Act and the popular politics leading up to the decision and for the indefinite future actually have little or nothing to do with existing healthcare that actually exists.  There are no death panels; universal coverage is not socialism; and no country outside of places like Somalia and the Eastern Congo believe that the government should not have a place in healthcare.

We live in a time when we’ll soon be spending one (1) out of every five (5) dollars of our GDP on healthcare.  Our employer-sponsored healthcare system makes no business sense and, as stated at the beginning of this article, fails so many who die and suffer needlessly.  Fragmentation and an ad hoc employer-based anti-system is expensive and leaves too many people without insurance, which ends up being a driver for even inflationary healthcare expenditures.

While conservatives bemoan the Supreme Court’s decision to uphold the Affordable Care Act  under Congress’s taxing power, they ought to look at how dysfunction in the system is far more taxing to the average person’s wallet.  Too often the call for “consumer-driven healthcare” becomes a euphemism for cost-shifting to consumers rather than improved quality and cost-effective care.  Even in systems with universal coverage, there are vibrant private sector health care businesses.  From a technological perspective, the government has a vital role in areas such as standards, core infrastructure development, and regulatory frameworks to protect everything from privacy to the security of your health data. The right mix of regulatory standards and frameworks can actually help businesses flourish.  We need a great deal of policy innovation in healthcare to keep pace with technological change. Innovation is not just about simplistic calls for cutting “big government”.

In support of this last point we are have already seen how creative government can support the growth of innovative companies in healthcare and also create jobs.  Health and Human Services (HHS) launched its Health Data Initiative two years ago to open up government health data to entrepreneurs and citizens.  This is health data our tax dollars pay for, but which was left in formats that rendered it not very user-friendly until the Initiative was launched.  An event, the Health Datapalooza, was launched to bring entrepreneurs, venture capitalists, and governments together.  The 2010 event had 48 participants.  In early June 2012, the most recent Datapalooza had over 240 companies and 1,500 attendees and was a dramatic demonstration of how government open data initiatives armed with an entrepreneurial focus can create jobs and new products and services that can have a big impact on health.  Polarized ideological posturing won’t solve the health crisis.

We also have significant technological challenges around the issue of interoperability.  Having the ability to choose different devices and technologies to collect and store our data and to be able to move that data from one health plan to another or across healthcare providers is no easy feat.  Most large healthcare IT programs in the 1990s–early 2000s were very capital intensive, lengthy technology implementations that are expensive to replace.  This has made many healthcare providers wary of new technologies.  One of the issues that both Republicans and Democrats agreed on was the incentive for electronic medical record (EMR) adoption in the stimulus package.  In theory this serves as the foundation for creating data liquidity in the future.  In reality, the practice of interoperability is challenging and is as much a human and organizational challenge as a technological one.  The right approach to interoperability is urgently needed and can help reduce uncertainty and risk for medical device manufacturers and even for those working down stream in the area of analytics.  At the Open mHealth Project (where I’m an advisory board member), we are working on filling in some of these gaps that challenge our ability to create personalized and optimized data ecosystems through an open source architecture.  We believe this will reduce the amount of effort and resources to develop devices and data analytics for evaluating and inferring the impact of interventions.

Data privacy, ethics, and ownership are emerging issues that we need to address as well. The emergence of Big Data allows us to mine and analyze massive amounts of data—both structured (e.g., traditional health services research) and unstructured (e.g., data from Twitter, Facebook, or much of the data in an EMR)—and has brought the issue of privacy to the fore.  It is now possible to de-anonymize data from a large national anonymized survey, for example.  The World Economic Forum recently issued a paper that argues that personal data has now become a new asset class.  Jerry Michalski, in addition, recounts some of the emerging ethical issues that Big Data raises – from the ways data can be used against us (a big issue in healthcare) to making us over-confident in projections and forecasts (as the financial crisis and numerous banking fiascos in recent months, e.g., J.P.Morgan, have demonstrated).  Yet there is much to be gained here from early outbreak detection to creating new efficiencies in health systems and better clinical decision-making with personalized health data.

Much work needs to be done in this area to make sure that the circle of beneficiaries of Big Data is wide and includes the marginalized.  I’ve been advocating the creation of a Global Data Alliance, a public-private partnership that can bring the best minds on data together to create the policy and ethical frameworks that build trust and value across the stakeholder value chain.  A key feature of the Alliance would be to foster public-private partnerships to address market failures in the data value chain, from food systems to healthcare systems and insurance.  Of the utmost importance are innovations that focus on products and services for low-income populations to mitigate their risk.  Risk mitigation, particularly in healthcare, should not be the preserve of the top one percent.  We need to build on Robert Kirkpatrick’s notion of “data philanthropy” for tangible social businesses in the healthcare space that are capable of improving health outcomes and changing risk equations for the most needy.

Last week’s Supreme Court decision is only the beginning of a long road to forge a healthier nation. We have the technological tools to create efficiencies and build transformative health systems. Now we need the human interoperability and creativity to realize the visions the technologists.  That will require political leadership and imagination that has become a scarce resource at a time when we need it most.

_____________

Dr. Jody Ranck has worked on health policy and health technology for over 20 years as a researcher and strategist.  Much of his work has been in global health working on programs in Africa and South Asia.  He has worked extensively with both private sector and government agencies in the area of technology forecasting and innovation strategy.  He recently served on a committee at the Institute of Medicine on the use of information technologies and is an analyst for GigaOM, which recently published his book, “Connected Health: How Mobiles, Cloud and Big Data Will Reinvent Healthcare”.

This article is available online at:
http://www.forbes.com/sites/benkerschberg/2012/07/06/how-connected-health-public-private-cooperation-and-big-data-can-revolutionize-health-care/