To most people, the outlook for nuclear power wouldn’t seem bright. The Fukushima disaster in Japan three years ago increased public resistance to the industry. Cheap natural gas is undercutting its competitiveness. Aging nuclear plants around the country, including Vermont Yankee in Vernon, Vt., are shutting down.

But into this bleak environment come two startups with roots at MIT hoping to revive an industry that has long struggled to make a comeback. Their technologies aim to solve issues that have bedeviled nuclear power for decades: safety, cost, and radioactive waste.

Transatomic Power, a three-person firm sharing incubator space at the Cambridge Innovation Center, is designing a reactor that would be cheaper than coal and generate electricity from spent fuel rods — aka radioactive waste — piling up in the nation’s nuclear plants. UPower Technologies is developing a miniature atomic power plant that would be cheaper and cleaner than diesel generators used in remote locations.

Despite the political, economic, and technical challenges facing the industry, these companies, and handful of other startups, are betting that the increasing urgency of climate change will mean a bigger role in the energy mix for emissions-free nuclear power.

Last week, in an alarming report known as the National Climate Assessment, a panel of scientists concluded that climate change, accelerated by the burning of oil, coal, and natural gas, is already having serious effects. Many parts of the nation, including the Northeast, are experiencing them in the form of violent storms, increased flooding, extended droughts, and severe wildfires.

“Traditional environmentalists might shudder at this,” said Richard Lester, head of MIT’s department of nuclear science and engineering, “but our students see themselves as able to respond effectively to the climate change threat by devising cheaper, safer nuclear-generated electricity. That’s what’s motivating them.”

Leslie Dewan and Mark Massie got the idea for Transatomic as students at MIT, where they studied nuclear engineering. They saw the potential for nuclear to supply large amounts of electricity without producing greenhouse gases that raise global temperatures.

The pair decided to focus on improving the safety of power plants by doing away with water. Fuel rods are submerged in water and heat from the uranium fuel is converted into steam to turn a turbine and generate electricity. After a few years, however, the metal assemblies that hold the rods must be removed, even though only about 5 percent of the energy content is used.

Nuclear plants also rely on water to cool reactors in case there’s a need for a rapid shutdown. In the Fukushima disaster, pumps that circulate water to the reactors lost power, leading to dangerous overheating and core meltdowns.

Instead of water, Transatomic’s design uses salt. Uranium is dissolved in a tank of molten salt, meaning there are no metal assemblies to damage. As a result, the fuel can stay in the core for decades until nearly all its energy is extracted.

This technology, Transatomic added, could recover the remaining energy from spent fuel rods stored in US nuclear plants — more than 70,000 metric tons, enough to cover a football field about seven yards deep.

With the molten salt approach, the plant could cool itself without pumps, according to Transatomic. The salt and fuel in the core would simply drain to a containment tank, slowing the heat-producing nuclear reaction and freezing solid in the tank within a few hours.

“This technology is so good and it’ll be so useful — that’s what makes me personally want to fight for this,” said Dewan, the chief science officer. “It can do a great deal of good for the world.”

Dewan and Massie launched Transatomic Power three years ago, and later signed Russ Wilcox, the cofounder of E Ink, as chief executive. Wilcox sold his company, a maker of paper-like displays for e-books, to a Taiwanese firm in 2009.

Wilcox, who saw Dewan and Massie present their innovations at a conference, said he was inspired by their vision of an energy technology that could reduce the stress of drilling, mining, pollution, and other human activities on the world’s natural resources. “It was a powerful moment,” he recalled.

A growing number of entrepreneurs say it’s a good time to innovate in nuclear power, in part because of advances in other fields. Supercomputers, for example, can simulate material and nuclear reactions, saving the time and expense of building experimental reactors.

The Energy Department, led by an MIT physicist, Ernest Moniz, supports nuclear innovation as well. In announcing a grant to advance small modular reactors — scaled-down versions of today’s plants — Moniz said in December that the Obama administration is committed to “strengthening nuclear energy’s continuing important role in America’s low-carbon future.”

As Wilcox shows, the possibilities for nuclear power are attracting successful entrepreneurs from other areas of technology. Bill Gates, the Microsoft founder, is bankrolling TerraPower, a company near Seattle that also is designing a reactor to run on spent nuclear fuel.

Nathan Gilliland was a cofounder and president of Waltham-based Harvest Power, which develops projects to convert food and other organic waste into fertilizer and energy.

Earlier this year, he signed on as chief executive of a Canadian nuclear startup, General Fusion, that is trying to use nuclear fusion, as the sun does, to produce energy.

“The amount of energy you can create from just a small amount of nuclear fuel is so dramatically different than renewables,” Gilliland said. “It could really help solve the planet’s energy needs.”

The energy needs of remote parts of the planet are the target of UPower Technologies, another MIT startup.

The Boston firm, launched last year, designed a nuclear “battery” small enough to fit in a two-car garage and provide several years of emissions-free energy to areas that rely on expensive — and dirty — diesel generators.

The smaller size also means smaller development costs — thousands of dollars to test a full-size prototype compared to millions for other nuclear power technologies, said Jacob DeWitte, chief executive of UPower and a PhD candidate at MIT.

The firm has won about a total of about $100,000 from business plan competitions, including the MIT Clean Energy Prize and the Boston accelerator program MassChallenge.

“I’ve always been inclined toward entrepreneurial activity,” DeWitte said, “and really thought that nuclear was underutilized.”

Still, introducing new nuclear technology, never mind transforming the industry, requires a long, uphill climb.

Nuclear faces opposition from environmental groups, which say the risks from waste, accidents, and proliferation of nuclear weapons remain unacceptable.

Edwin Lyman, a senior scientist at the Union of Concerned Scientists, a science-based advocacy group in Cambridge, said even the latest safety improvements can’t take into account all possible catastrophes, such as the tsunami that devastated the Fukushima plant.

Nuclear startups also face lengthy regulatory processes and development costs that could reach into the billions. Transatomic so far has raised $1 million, but said it needs to raise $100 million or more to finance a scaled-down prototype reactor. It estimates that building its first full-size power plant could take 15 to 25 years.

Finding investors willing to put up this kind of money and wait that long for a potential payoff is another daunting challenge, industry officials said. More government funding of research and development could help, said Ray Rothrock, a venture capitalist at Venrock in Palo Alto, Calif., but regulations also need to be overhauled to accommodate new technologies, particularly a faster licensing process.

“The regulatory framework has to be able to adapt to all sorts of innovations,” said Rothrock, who invested his own money in Transatomic. “And right now, it can’t — it’s simply not equipped for it.”