After years of making financial plans at huge companies like Unilever and Electronic Arts, Mat Fogarty realized that management is often out of touch with what is going on in a business. It isn’t entirely the executives’ fault: sometimes they don’t have enough information. Too many realities about how their companies are doing—things the executives have to know to make smart decisions—get bottled up at lower levels of the business. For example, several years ago, employees down the chain at Electronic Arts knew that a certain video game was unlikely to ship on time. But executives continued to promise an overoptimistic release date to the outside world, and then they spent millions in a frantic race to meet the deadline. It was too late: the game came out at least a year later. And even though it proved successful, the delay hurt the company’s earnings and its stock price.

In 2007, Fogarty left EA to found a company that could tap the expertise and insights collectively held by rank-and-file employees. Before long he had a software application that lets businesses set up a “prediction market.” It asks employees to bet, in play money, on what they think will happen. “When will this product really be ready to ship?” they might be asked, or “How many units will we sell?” The aggregation of responses can produce remarkably accurate answers and improve the way companies operate. His first customer was Electronic Arts.

Fogarty’s company, Crowdcast, is a newcomer on the second annual TR50, our list of the world’s most innovative technology companies. We look for companies that are setting the agenda in an increasingly important market, on the verge of disrupting an established market, or creating an entirely new market. What does it mean to set the agenda? It doesn’t necessarily equate to having the biggest market share, or else Intel and Microsoft would be in the TR50. Instead, we look for companies whose ideas and technologies are being imitated by other companies.

Timing plays a huge role in our choices, because technology markets change quickly. Last year we picked Athenahealth for its online health records system, a technology that was getting a push from government policy and other factors. But different innovations in biomedicine now appear ready to have more impact. For instance, researchers have made impressive progress in developing treatments using stem cells. That’s why we’ve added two stem-cell companies, Geron and Cellular Dynamics International, to the list. Another new TR50 company in the right place at the right time is ­BrightSource Energy, a provider of solar thermal power. The technology uses large arrays of mirrors to focus sunlight on a tower, generating heat that is used to produce electricity. BrightSource is taking advantage of federal funding for alternative energy (including a $1.4 billion loan guarantee) and state mandates requiring California utilities to use more green power. It’s building what will be, at least for a time, the world’s largest solar plant, a 392-megawatt facility in the Mojave Desert.

Explore the TR50 list. Browse by topic: Energy, Transportation, Computing and communications, Web and digital media, Materials, and Biomedicine.

Some companies have fallen off the TR50 because they didn’t find an innovative solution to new challenges. Adobe Systems struggled to deal with Apple’s refusal to make iPads and iPhones capable of running Adobe’s Flash software. Plastic Logic is gone from the TR50 now that its Que e-reader, which promised to be especially lightweight yet durable, has been canceled. When we picked Hulu last year, we said it was at the forefront of putting premium video content on the Web. But that description better fits Netflix now. Hulu, which is majority-owned by the parents of three of the big four U.S. broadcast TV networks, has some advantages, including the ability to let people stream shows even before their season is over. But it also has stopped offering everything for free and tried to sell a subscription package. Meanwhile, Netflix has cleverly woven streaming TV shows and movies into its existing DVD subscription services.

Other companies are joining the TR50 because their technologies are opening new markets. As gene-sequencing technology evolves, for example, the price of analyzing DNA is plummeting, and several companies have machines for sale. Last year Life Technologies bought Ion Torrent, whose founder, Jonathan Rothberg (see Q&A), has chosen to make a sequencing machine that is much cheaper—albeit less powerful for now—than his competitors’, which are sold mostly to research labs. Rothberg wants to create a market for gene testing in doctors’ offices and other clinical settings. By quickly analyzing certain segments of a cancer patient’s DNA, for instance, a doctor could better assess potential treatments. Another physician could get fast insights by putting a sample of blood through a machine offered by Claros Diagnostics, another addition to the TR50. With Claros’s microfluidic technology, a liquid is pushed through tiny channels on a chip that can analyze such anomalies as the elevated protein levels that can be a sign of prostate cancer.

Selecting the TR50 isn’t simple, but some companies are easy picks because their technologies jump out as fresh ways of doing things. Lyric Semiconductor has redesigned the microprocessor so computers can better deal with probabilities; such an approach could make fraud detection faster and recommendation software smarter. Or check out the “augmented reality” software from a Dutch company called Layar: it fills the screens of mobile phones with information about the user’s real-world surroundings. PrimeSense, based in Israel, developed technology that lets people play video games without a controller. It uses an infrared projector and camera and a special chip to detect movement in three dimensions, so players can manipulate the on-screen action with gestures and body movements. It’s available in Microsoft’s Kinect unit for the Xbox 360, but PrimeSense’s technology could also be used to control TVs and computers. That makes it a breakthrough with wide potential applications—a great definition of innovation.

Brian Bergstein is deputy editor of Technology Review.