Fifty years of DARPA: Hits, misses and ones to watch

Founded to protect the US against 'technological surprise', the agency has achieved some spectacular successes - and failures - in its 50-year history

SUCCESSFUL PROJECTS:

The internet: Precisely who 'invented' the mass of linked computer networks that is today's internet is a moot point. But it wouldn't have happened without the ARPANET network built by DARPA in the 1960s. The idea was to make a "self-healing" communications network that still worked when parts of it were destroyed. It was the first network to transmit data in discrete chunks, not constant streams, and led to the development of the TCP/IP specification still in use today.

GPS: We would be quite literally lost without today's global positioning system (GPS). But long before the current NAVSTAR GPS satellites were launched, came a constellation of just five DARPA satellites called Transit. First operational in 1960, they gave US Navy ships hourly location fixes as accurate as 200 metres.

Speech translation: Although not yet available to consumers, handheld language translation devices developed with DARPA funding are already being used in Iraq. Although accuracy can be as low as 50%, the devices have met with favourable reviews from forces on the ground.

Stealth Planes: It's probably the best example of DARPA fulfilling its remit to come up with "surprise" technologies - even the US Air Force was surprised by the idea. The first prototype, Have Blue, was tested in the late 1970s and became the precursor to F-117 Nighthawk stealth fighter.

Gallium Arsenide: One of DARPA's lesser known accomplishments, semiconductor gallium arsenide received a push from a $600-million computer research program in the mid-1980s. Although more costly than silicon, the material has become central to wireless communications chips in everything from cellphones to satellites, thanks to its high electron mobility, which lets it work at higher frequencies.

FAILED PROJECTS:
Hafnium bombs: In an episode reminiscent of the Cold Fusion debacle, DARPA forked out $7 million in the 1990s for research into a bomb predicted to release huge gamma-ray bursts without creating any nuclear fallout.

The theory was that hitting a small amount of a radioactive isomer of the super-expensive metal hafnium with X-rays would release this torrent of energy. No proof this could happen was ever found.

The mechanical elephant: Frustrated by a lack of decent tarmac in the jungle, DARPA sought to create a "mechanical elephant" during the Vietnam war. Its vision of high-tech Hannibal's piloting them through the forest never came true. It is alleged that when the director heard of the plan he scrapped the "damn fool" project immediately in the hope no one would hear about it.

Telepathic spies: One of the agency's most infamous blunders was its 1970s psychic spy program, inspired by reports that the Soviets were researching the area. DARPA invested millions to see if telepaths and psychokinetics - who claim to move objects using thought alone - could carry out remote espionage. They couldn't.

FutureMap: This program hoped to use a kind of terrorism futures market to predict key developments and even attacks. It was thought market valuations of possible future events could reflect the probability of their occurring. However, FutureMap was scrapped in 2003 after the notion of betting on terrorist atrocities was called "ridiculous and grotesque" by US politicians.

Orion: Set in motion shortly after DARPA was created, Project Orion aimed to drive an interplanetary spacecraft by periodically dropping nuclear bombs out of its rear end.

The entire craft was designed like a giant shock absorber with the back covered in thick shielding to protect human passengers. Concerns about nuclear fallout and the signing of the Partial Test Ban Treaty ended the project in the early 1960s.

ONES TO WATCH:

Robot Cars: DARPA's Grand Challenge competitions have aimed to foster the creation of driverless cars capable of travelling long distances across difficult terrain and even through busy traffic.

Thanks to the competitions, some impressive vehicles have already been produced and surely it won't be long before DARPA's robot cars are used in real military or civilian scenarios.

Z-man: The aim: to allow soldiers to scale vertical walls without ropes or ladders at a rate of 0.5 metres a second. The solution: mimic the microscopic hairs, or "setae", that allow geckos to stroll up walls and across ceilings. Small robots that climb using synthetic setae have already been demonstratedMovie Camera, but DARPA hopes to extend this technology to humans.

Underwater Express: Troop-transporting torpedoes could travel at speeds of up to 100 knots thanks to a phenomenon called supercavitation. This occurs when an object moves fast enough to vaporise the water around it into a single enveloping bubble.

With virtually no contact between the torpedo and the water, drag is reduced by up to 70%. But tests so far have been restricted to un-manned drones.

Bionic Limbs: DARPA wants prosthetic limbs that are "fully functional, neurologically controlled and have normal sensory capabilities" and is funding scientists who are making serious progress.

For example, Video of a bionic arm built by the creator of the Segway shows impressive dexterity, while other teams have built prototype prosthetics controlled by thought alone.

Switchblade: DARPA has already revolutionised aerial warfare by encouraging the development of unmanned aerial vehicles (UAVs). The Switchblade should go further by becoming the first supersonic UAV capable of long-range missions.

The $10.3-million project involves a bizarre design. The aircraft's 61-metre wingspan pivots by 60° as it gains speed, until one wing points forwards and the other backwards (see image) enabling it to reach speeds of Mach 2.

Exoskeletons: Think of an exoskeleton as a mechanical full-body upgrade. DARPA wants to increase the distance soldiers can march and the loads that they can carry with them.

An exoskeleton developed by MIT with DARPA funding recently proved it could make a 36-kg load feel 80% lighter.