An aircraft that can take off and land vertically like a helicopter—and yet cruise fast and far like an airplane—is arguably the holy grail of U.S. military aviation.

After decades of near misses, the Pentagon’s scientists think the hybrid-flight grail is finally within reach. In early March, the fringe-science Defense Advanced Research Projects Agency announced it was awarding a contract to Aurora Flight Sciences, a Virginia aerospace firm, to build a new kind of vertical-takeoff and -landing, or VTOL, aircraft—one that the agency claims can overcome fundamental problems with hybrid aircraft designs.

Aurora calls the new experimental aircraft “LightningStrike.” In its prototype form, LightningStrike will be on the small side—weighing in at just five or six tons—and robotic. The military and Aurora both said that it would be possible to scale up LightningStrike and add human pilots.

The purported secret to safe, efficient hybrid flight is in the craft's wing. Rather than giving the vehicle fixed wings and making the engines or the engine nozzles rotate in order to switch between vertical and forward flight—that's the typical VTOL layout—Aurora's engineers are combining wing and engines into a single unit and making the whole thing rotate.

“Just when we thought it had all been done before, the Aurora team found room for invention—truly new elements of engineering and technology that show enormous promise,” Ashish Bagai, DARPA’s VTOL program manager, said in an agency statement.

But there are reasons to be skeptical. Aurora's LightningStrike design actually has more in common with previous, failed hybrid aircraft than the company or the government would probably like to admit.

To be sure, the armed forces are serious about developing a dependable VTOL craft, so their optimism is perhaps understandable. Commanders desperately want the flexiblity of a helicopter, but without a helicopter's traditional liabilities in speed and range. And they also crave the comparative swiftness and endurance of a fixed-wing airplane, but would like to shake off the plane's reliance on big, expensive, vulnerable runways.

But every technology the military has developed to combine the best attributes of a helicopter and an airplane has resulted in a machine that is complex, unreliable and even dangerous to fly—something DARPA's readily acknowledged.

"For decades, aircraft designers seeking to improve vertical-takeoff and -landing capabilities have endured a substantial set of interrelated challenges," the research agency said in a statement. "Dozens of attempts have been made to increase top speed without sacrificing range, efficiency or the ability to do useful work, with each effort struggling or failing in one way or another."

Take for example, the AV-8B Harrier, a British-designed jet fighter dating to the 1960s that can launch and land vertically thanks to its rotating engine nozzles. Technically, the Harrier checks the most important boxes for an effective VTOL warplane. It doesn't need a long runway. It can accelerate to nearly the speed of sound.

But the Harrier—which today equips several U.S. Marine Corps squadrons—is mechanically complex and, as a result, unreliable and wickedly unsafe. No fewer than one-third of all the Harriers in U.S. service have crashed, killing scores of pilots and making the jet one of the most accident-prone in the modern military inventory.

The V-22 Osprey, a hybrid "tiltrotor" built by Boeing and Bell that launches vertically but cruises like a plane thanks to its rotating wingtip engine nacelles, is only slighlty less unsafe. Dozens of service members have died in V-22 crashes since the tiltrotor began flying in the late 1980s.

Most of the crashes have been caused by the tricky physics involved in hybrid flight, the tendency of any low-flying aircraft to kick up blinding dust clouds or by a related problem—a complex, fragile engine design that simply can't keep up with the high demands of VTOL operations and tends to fail when exposed to dust.

And here's the kicker: if a V-22 suffers an engine failure, it can't glide like a fixed-wing airplane can or “float” to the ground like a helicopter can thanks to the tendency of rotors to keep spinning—a.k.a., “autorotate”—even after the attached engine flames out. The Osprey is aerodynamically ungainly and, if an engine fails, tends to fall to the ground like a rock.

But the Marines and Air Force have bought hundreds of Osprey, anyway—and the Navy is getting ready to acquire some, too. To protect the program from cancellation, the military has made a concerted effort to shift blame for accidents away from the V-22. In several cases where Ospreys tumbled out of the sky due to obvious mechanical failure, government investigators still officially blamed the pilots for the crashes.

Bagai told The Daily Beast that LightningStrike should sidestep many of the V-22’s worst flaws. When landing quickly, an Osprey’s two big rotors can create turbulent air, potentially causing the craft to spin out of control. That's called “vortex ring state,” and LightningStrike is immune to it because of the aerodynamics of its many, small rotors, Bagai said.

But in solving the vortex-ring state problem, LightningStrike could risk entirely different problems. Among them, a kind of low-altitude turbulence unique to helicopters and VTOL craft. "Wake-induced instabilities, when hovering in proximity to the ground, for instance, are more likely to upset the aircraft," Bagai told The Daily Beast. The solution, the DARPA scientist said, is to program LightningStrike's control computers to compensate for the rough air.

But there is no easy way to deal with the dust that has choked so many low-flying V-22s and other rotorcraft—and which could also endanger LightningStrike. DARPA has asked Aurora to do something—anything—to help LightningStrike manage dust clouds. But swirling debris clouds are inevitable as the LightningStrike comes in for a landing, Bagai said. "Jet-like flow-fields produced by the current design will, no doubt, entrain dust and loose material."

Nor is there an easy solution to the problem of what to do if the LightningStrike loses power. True, the drone could survive and keep flying if just a few of its rotors got knocked out by a mechnical problem or an errant bird, Chip Sheller, an Aurora spokesman, told The Daily Beast. “The highly redundant nature of the electric distributed propulsion system allows us to maintain controlled flight should the aircraft lose one or more fans to equipment failure or bird ingestion.”

But if the single Rolls Royce AE1107C turboshaft engine that powers the rotors were to fail, all the rotors would stop spinning and the LightningStrike could turn into a flying brick. “Autorotation with hybrid designs such as this is not physically possible,” Bagai said. "Other means of controlled landings in the event of full power-out would need to be developed and verified."

Perhaps most worryingly, the LightningStrike’s AE1107 engine is the same one that powers the V-22—and has proved unreliable. The Marine Corps expected each engine to run for up to 600 hours before needing an overhaul, but in 2011 Sen. John McCain, an Arizona Republican, noted that the AE1107s were lasting just 200 hours.

The associated maintenance cost was “eating up the Marine Corps’ budget,” McCain said. If the military decides it likes LightningStrike and modifies it for frontline use but keeps the same engine, it could could end up paying the same high cost for its supposedly new-and-improved VTOL craft.

DARPA’s not worried ... yet. “As an experimental aircraft, these are the types of challenges that such a program is intended to encounter,” Bagai said. "We can't expressly mitigate every anticipated issue for production applications at this stage."

To be fair, no one ever promised that the search for the aviation holy grail would be an easy one. Bagai said that DARPA and Aurora would deal with LightningStrike’s problems as they arise.

That’s typical of DARPA’s try-and-see approach to developing technology. New equipment is bound to have issues. “We do our best to ensure they aren’t ‘fatal,’” Bagai said.