The challenges are massive—recharging the battery is tricky, and durability is an issue. Also, to get the battery to release power quickly enough for passing requires expensive nanotechnology. “It remains a long shot, but we’re excited by the potential,” says Winfried Wilcke, IBM’s lead researcher on the project. Big Blue sees a prototype by 2013 and commercialization as early as decade’s end.

2. Harnessing the sun’s power

Ever since cold fusion flopped spectacularly, the idea of finding an affordable way of replicating the sun’s method of generating energy has become almost a joke. That may be about to change. Yes, the two major fusion reactor designs being explored in the research world—one is called a tokamak and the other is inertial confinement systems—show promise, but they are 20 to 30 years off. Also, they require either gigantic superconducting magnet systems or extra-fierce laser arrays and will cost tens of billions at best.

A small Canadian company called General Fusion has a new technology called magnetized target fusion, which could end up costing a fraction of what the other designs do. The privately funded project has attracted $32 million from investors including Amazon CEO Jeff Bezos. This is not the way things are traditionally done in Big Science, where only governments with deep pockets can afford to fund the R&D and construction of huge experimental machines. “If we had a 20-year project, we would never get the money. But we’re promising to demonstrate a net energy gain in five years, and we’re meeting our milestones,” says CEO Doug Richmond.

General Fusion’s design shares aspects of the other two but aims to pull off fusion reactions using far simpler and cheaper hardware. The company’s reactor draws on work begun more than 30 years ago at the U.S. Naval Research Lab. It uses 200 very powerful pneumatic pistons inserted through holes in a spherical metal vessel that Richardson compares to a “wiffle ball.” Firing the pistons at precisely the right moment creates an acoustic shock wave that compresses hot liquid lead and lithium swirling around inside. Then, the injected deuterium-tritium gas encounters such heat and pressure that a fusion reaction briefly occurs. General Fusion brought the Navy’s original idea to life by engineering control systems for the pistons that weren’t possible a few decades ago.

If this long-shot technology works, it could provide clean, safe energy for generations to come.

3. Creating electricity in space

The idea of beaming solar power down to Earth from space was popularized in a 1941 Isaac Asimov short story in which the machinery was controlled by a robot called Cutie. Today, solar space stations still sound far-fetched, but scientists in the U.S. and Japan are pursuing modern versions of the system, which are becoming more feasible as space flight and solar panels promise to become more affordable.

How would it work? The panels would orbit in space—immune from rain, clouds, and nighttime—gathering solar energy 24/7. The panels would be 43 times more efficient than land-based ones, says Col. M.V. “Coyote” Smith, who has studied the concept for NASA. The satellites would then beam the energy to Earth in the form of microwave radiation. Implausible? John Mankins, the former head of advanced concept studies at NASA, has conducted successful tests in Hawaii, sending wireless electricity between two islands.

The hang-up is cost. Building a big space solar operation would cost billions, Mankins says. While a couple of universities are working on it, skeptics abound. “If a potential investor sat down and penciled out the costs, they would stop returning your phone calls,” says Seth Masia, editor of Solar Today magazine. Still, new projects like Microsoft co-founder Paul Allen’s aircraft, which one day could affordably launch satellites, and the fact that solar panels are getting cheaper, are making this technology suddenly seem more science than fiction.

4. Tapping fuel from the ocean floor

In the race to find an economically viable biofuel, researchers are now looking at a surprising source: seaweed. While making cheap fuel from pond algae has proved difficult, the potential advantages of seaweed, or macro-algae, are big. It’s one of the world’s fastest-growing plants, doesn’t need fertilizer, requires less acreage than land-based crops (plus, no clear-cutting to make way for farms), and its fuel would emit less CO2 than the current ethanol champion, corn.

More important, more than half the dry mass in seaweed is sugar, which is the new crude, fuel scientists say. That’s because sugar can be easily converted to ethanol or butanol. Bio Architecture Lab, a startup, has partnered with DuPont (DD), the Department of Energy’s ARPA-E labs, and the venture arm of Norway’s Statoil (STO) to develop the chemistry that would unlock the energy in that sugar and create a fuel that’s cheaper than the alternatives.

Bio Architecture Lab, headquartered in Berkeley, has built three seaweed farms off the coast of Chile. Workers using winches and lines harvest giant strands of seaweed from boats. The company recently broke ground on a pilot ethanol manufacturing plant in the Los Lagos region of Chile, slated to start operations next year. The challenge: It’s one thing to make small amounts of fuel in a lab, but fuel production is a big global business. Will this technology be able to scale affordably?

This article is from the January 16, 2012 issue of Fortune.