Thinking About the Future: An Introduction to Forecasting

From the third chapter of "Thinking about the Future," a book co-edited by Social Technologies' Andy Hines and futurist Peter Bishop, comes this Introduction to Forecasting.
Forecasting involves creating alternative futures. Most organizations, if not challenged, tend to believe the future is going to be pretty much like the past. If the analyst probes at an organization's view of the future, he or she will find an array of unexamined assumptions, which tend to converge around incremental changes--a.k.a. the "official future" or "baseline forecast"--that pretty much preserves the current paradigm or way of doing things. A key task for the analyst, therefore, is to challenge this view and prod the organization to take seriously the possibility that things may not continue as they have--in practice, they rarely do!
In essence, forecasting involves generating the widest range of creative possibilities, then consolidating and prioritizing the most useful for the organization to actively consider or prepare for as it moves forward.
A principal means of challenging the official future is to develop alternative futures. A key tenet of strategic foresight is that the future is inherently unknowable and efforts to get it exactly right are futile. What the analyst can offer is to expand the range and depth of possibilities for the organization to consider, thereby reducing the likelihood and magnitude of surprise. This in turn enables the organization to successfully navigate through what does emerge. The analyst may be asked to produce the "correct" future, that is, to predict what will happen. Organizations prefer the clarity of dealing with a single possibility rather than the messiness of alternatives. In response, the analyst must get across the idea that single-point forecasting is doomed and the organization will be better served by understanding and preparing for a range of possibilities.
Forecasting alternative futures does not mean developing detailed plans for every contingency, however. Rather it means monitoring the external environment for leading indicators--signs or guideposts that suggest events are heading towards one or more of the alternatives.
The guidelines in this section suggest how the organization can produce a useful set of alternative futures. A useful forecast challenges existing assumptions about the future, gets the organization to consider "what if," and thereby motivates it to plan and act differently. An alternative future that turns out to be off-base can still be useful if it has prompted the organization to take the future seriously, to consider and prepare for possibilities in a way that yields helpful learning and experience.
The preparatory work done in Framing begins to pay dividends here, incidentally. If Framing has helped the team function well, produced a work environment that is conducive to creativity, and so forth, the work done in Forecasting will be that much better.

Germany's CESAR crowned king of rovers in ESA’s Robotics Challenge

A robot rover designed by a Bremen university team has won an ESA contest to retrieve soil samples from a lunar-style terrestrial crater. Eight student teams fielded rovers during the event, their progress monitored by an advanced 3-D viewer already flight-tested in space and planned for eventual deployment on the Moon.
Craters surrounding the Moon's poles are a top 21st Century science target. Lunar researchers believe these craters may be 'cold traps', preserving ancient water ice deposits. Such ice would not only be an invaluable time capsule, it would also support manned lunar settlements. But the only way to verify the ice is there is to go fetch it, which is where rovers come in.
The bleak pumice landscape of Minas de San José within Tenerife's Teide National Park stood in for the Moon during the inaugural ESA Lunar Robotics Challenge (LRC). Built within strict size, weight and power constraints, the rovers had to descend down the steep 40° slopes of a 15-metre deep crater, grab 0.1 kg of specifically selected soil then carry it out again – all the while in darkness.
Working from a trailer camp 2000 metres up, each five-strong team was confronted with some distinctly non-lunar weather including heavy rain and clouds. In the event only one rover managed to complete the assignment –Bremen's three-wheeled CESAR (Crater Exploration and Sample Return) robot, duly judged LRC winner on 26 October.
The rovers were not the only hardware undergoing field testing. "The location was an excellent place to exercise the capabilities of the Erasmus Recording Binocular (ERB)," explained Massimo Sabbatini of ESA's Human Spaceflight Directorate. "This innovative high-resolution stereo-video recorder will become part of the standard toolset of astronauts when surveying lunar areas."
Despite being more than 2000 km away from the European mainland the test site was kept linked with the internet by a satellite ground station provided by ESA's Telecom Directorate, which enabled remote team members to follow the event and even debug rover software as needed.
Francesco Feliciani of ESA’s Telecom Directorate said: "The successful live transmission of the videos streams from the LRC proves that the technology is ready to be used in real life applications.”
The broadband-quality transmission was made possible by the ESA co-funded AmerHis regenerative payload hosted aboard Hispatsat’s Amazonas spacecraft. AmerHis works as an internet router in the sky.
Stefano Badessi also of ESA Telecom commented: “Our satellite terminal allowed those team members who could not be on the Teide in person to watch how their teammates were faring.”
Jury member Richard Fisackerly paid particular attention to the control stations teams employed to communicate with their rovers. As part of ESA's Aurora Programme developing technology for manned and robotic exploration of the Moon and Mars, Fisasckerly plans to apply the student experience into terminals planned for lunar astronauts to interact with robot assistants.
Adverse weather conditions affected preliminary trials and also meant two of the rovers – fielded by the University of Pisa and Scuola Superiore Sant'Anna – were forced to perform the Challenge in daytime rather than at night.
This being so, the jury decided against formally ranking the teams but did compile a scoreboard to give feedback on strategic choices. LRC judge and organiser Gianfranco Visentin said: "We hope that this information will increase the wisdom of the teams for future competitions."
Also affecting scoreboard figures, the University of Surrey's SELENE proved unable to take part due to a mechanical failure while Scuola Superiore Sant'Anna's pESApod was shorted out by rainfall.
Lucio Scolamiero, member of the jury and co-proposer of the LRC, declared “Only six months ago most of the rovers were only at drawing board level. What the student teams have done in such a short time has been a challenge in itself, they all deserve to be considered winners of this challenge”.
The eight rover teams shared their experiences on 14 November at a special session concluding ESA's 10th Workshop on Advanced Space Technologies for Robotics and Automation.

GERMAY'S ROBOTS

Friendly robots at Germany's K2007 exhibition
24 September 2007
var object = SHARETHIS.addEntry({title:'share',summary: 'Sharing is good for the soul.'},{button:false});
document.write('');
var element = document.getElementById("share");
object.attachButton(element);

Visitors to the K2007 event in Dusseldorf will be able to shake hands with one of ABB's IRB140 robots.The event will take place on 24-31 October 2007, and will give visitors literally a first hand inspection of the multitasking capabilities of robotic automation in the plastics industry. They will be able to lead an IRB140 robot by the hand and teach it any moves they like. This lead by hand technique lets the user hold the robot gripper while a command de-energizes the robot and it goes limp. The user then moves the robot by hand and simply demonstrates the required positions for a machine cycle or for part processing. In the case of part processing, such as cutting or polishing, the program afterwards re-runs the robot along the taught positions and automatically adds all necessary positions to generate the perfect processing path. The IRB140 robot is a compact and powerful 6-axis machine that combines fast acceleration, large working area and high payload. As an adjunct to user convenience, ABB introduces a unique new service for monitoring and fault finding in robot applications, whereby diagnostics and even predictive machine condition monitoring can be effected over the Internet. The robot controllers communicate wirelessly via an embedded web server to alert of service requirements – maximising uptime, preventing unplanned stoppages and minimising maintenance costs.

Robot Versus Sniper

Securing an intersection is a basic combat task that the Pentagon hopes will one day be tackled by unattended robots. In the scenario below, set in 2020, an armed robot has forged ahead of a squad (not shown) to determine if a sniper is stationed at a key corner. As simple as the mission might seem, it’s a huge engineering challenge to program the skills needed for the assignment into a robot’s brain. Experts say success will require integrated sensors to double for human sensory organs and powerful processing of the data to mimic human training—and instinct. “What is intuition?” asks Jon Bornstein, head of the Army Research Lab’s robotics office. “A series of cues that give a high probability of something occurring.”

The MULE

This prototype of the Gladiator, built by Carnegie Mellon, is equipped with smoke generators and grenade launchers. Marine Corps officials say they will field-test the latest, 1-ton version this year. (Photograph by Craig Cameron Olsen)
Despite the challenges, armed UGV development is on the rise. Foster-Miller is currently working on a successor to the SWORDS, a larger and more versatile robot called the MAARS (Modular Advanced Armed Robotic System). Technicians in the field will be able to replace the system’s M240 machine gun (the same kind currently planned for the MULE) with an arm or trade the tracks for wheels. However, the MAARS requires a human operator to move and acquire targets. IRobot, the maker of thousands of bomb-defusing PackBots, plans to introduce its Warrior X700 this year. The Warrior is larger than the PackBot and has a similar set of articulated tracks that allows it to climb stairs, and a 150-pound carrying capacity. The company is touting the Warrior’s ability to fight fires, haul wounded and serve as a weapons platform. But according to Joe Dyer, the president of iRobot Government & Industrial Robots division, a key benefit of an armed UGV isn’t what it can dish out, but what it can take: “A robot can shoot second.” The Warrior is able to follow GPS waypoints, can breach ditches and navigate cramped conditions on its own, but it will still rely heavily on human guidance in a fight. Where weapons are involved, Dyer says, “Autonomy’s going to come into robots on little cat’s feet.” Like their bomb-poking forebears, weaponized bots are disposable, making them particularly useful in urban warfare, with its high potential for collateral damage and sudden, point-blank firefights. “Robots are fearless, so there’s an opportunity to better assess the situation,” Dyer says. “That means less risk to noncombatants and to friendly forces.” In urban warfare, where troops often lose the high-tech edge, an armed ground robot is the perfect point man. “Send a robotic platform into a room, and it might take some small arms fire,” Shoop says. “But it can be repaired fairly easily. A soldier or Marine is not as easily repaired.” The MULE is toying with my emotions. After running through its full range of articulated positions—a hilarious diagnostic dance routine that has it pivoting, rising and tipping its wheels off the road—the robot is now ramming a car. The sedan offers little resistance, sliding across the asphalt. Like proud owners watching their pit bull tear through a chew toy, the small crowd of defense contractors and engineers are chuckling. Next, the MULE climbs onto a 5-ft.-high platform and prepares to cross a 6-ft. gap. The robot reels back on its hind legs. It inches forward and falls across the space, its front wheels slamming onto the next platform. Although it was moved into position by a human operator, the robot’s terrain-clearing performance was automatic, using internal sensors that track wheel position and speed and two onboard Pentium III processors cycling an array of mobility algorithms. Despite being blind, the MULE is already surprisingly autonomous. The exact details haven’t been worked out, but the goal is for a single sergeant to handle multiple robots. But no matter how sophisticated the robot, Lockheed officials point out, it will never fire without a command from a human operator. Having a person decide when to shoot is a recurring theme in discussions about armed robots. Maj. David Byers, the assistant program manager for the MULE, compares the likelihood of the robot’s weapons discharging accidentally to a modern tank inexplicably firing off a round. Using the UGV’s sensors, a human will confirm that each target is a hostile before firing. “Armed robots are still foreign to Army culture,” he says. “We need to cultivate the understanding that they are quite safe.” The demonstration is winding down, and after a slew of caveats and reassurances, it’s my turn to drive. On a grassy slope overlooking the track, an engineer hands me the Xbox 360 controller. I will not, I’m told, get to wear the shiny Rocketeer backpack. The game controller is surprisingly standard-issue—no external tweaks or mods. When I hit a button, the prototype rumbles forward. I jam the thumbstick to one side, and the robot turns in place, its fresh wheel screeching, painting a perfect circle on the asphalt. I guide the MULE through a small parking lot, around cars, and across a muddy patch to give the tires a little more traction. The robot is responsive, literally leaning into turns and braking with finesse. My fingers keep hitting the unused buttons, automatically probing for the one that opens fire. In the sci-fi cult classic The Last Starfighter, the teen hero is drafted into a galactic war. The arcade game he spent hours mastering was really an alien simulator, and with a quick costume change, he’s reborn as an ace pilot. For 10 minutes, my fantasy is much better: Years of Saturday afternoons and missed classes and so-called sick days spent clicking away at a game console—it wasn’t wasted time; it was training. I have become a crack military robot pilot. Time’s up, and I hand back the controller, the prototype still rumbling away, slightly muddier than I found it. We head down the hill, and as I pass an engineer, I mention how easy it is to drive. “Yeah, we based the controls on Project Gotham Racing,” he says. It’s a joke, but the quip offers a glimpse of what future warfare might look like—robotic, autonomous and just a little bit chilling.

America's Robot Army: Are Unmanned Fighters Ready for Combat?

At a muddy test track in Grand Prairie, Texas, 13 miles west of Dallas, the robot is winning. It has climbed on top of a sedan, its 2.5-ton bulk propped on the crumpled roof. The car never stood a chance.
digg_url = 'http://digg.com/tech_news/Meet_America_s_new_armed_robots';
The MULE (Multifunction Utility/Logistics and Equipment) is roughly the size of a Humvee, but it has a trick worthy of monster truck rallies. Each of its six wheels is mounted on an articulated leg, allowing the robot to clamber up obstacles that other cars would simply bump against. Right now, it’s slowly extricating itself from the caved-in roof, undulating slightly as it settles into a neutral stance on the asphalt. This prototype’s movements are precise, menacing and slow. When the final product rolls onto the battlefield in six years, it will clear obstacles in stride, advancing without hesitation. And, like the robot cars that raced through city streets in last fall’s Pentagon-funded DARPA Urban Challenge, the MULE will use sensors and GPS coordinates to pick its way through a battlefield. If a target is detected, the machine will calculate its own firing solutions and wait for a remote human operator to pull the trigger. The age of killer robots is upon us. But here at defense contractor Lockheed Martin’s test track, during a demonstration for Popular Mechanics, this futuristic forerunner of the robot army has a flat tire. “Actually, this is good,” says Michael Norman, Lockheed’s project manager for the prototype. “You’ll be able to see how quick it is to swap in a new tire.” He nods toward an engineer holding an Xbox 360 controller and wearing a gigantic, gleaming backpack that contains a processing computer. The engineer taps a handheld touchscreen. One of the robot’s wheeled legs rotates upward, and a two-man crew goes to work. Each leg has its own hub motor to allow for a variety of ­positions. If one leg is blown off by enemy fire or a roadside bomb, the rest are able to soldier on, with the robot automatically adjusting its center of gravity to stay mobile. It’s highly functional. But with its engine powered down—it runs on a Mercedes-built engine originally modified for unmanned aerial vehicles (UAVs)—and one leg cocked gamely in the air, the MULE doesn’t look so tough right now. In fact, the MULE isn’t ready for battle. Barely a year old, the prototype is a product of the Army’s Unmanned Ground Vehicle program, which began in 2001. It has yet to fire a single bullet or missile, or even be fitted with a weapon. Here at the test track it’s loaded down with rucksacks and boxes, two squads’ worth of equipment. At the moment, the MULE has no external sensors. “We’re 80 percent through the initial phase,” Norman says, “but we don’t have the perception fully tested. It knows heading and speed, but it’s blind.” In other words, it’s essentially one of the world’s biggest radio-control cars. And, eyeing the robot’s familiar controller, I realize I might have a shot at driving it. I know my way around a video-game console, but the engineers are noncommittal about my request to drive the MULE. The goal, of course, is for the MULE to drive itself. Sitting a short distance away is the prototype’s future, a full-size mockup of a weaponized variant, its forward-facing machine gun bracketed by missile tubes. The gleaming sphere set on a short mast looks precisely like a robot’s eyeball. It will visually track moving targets, allowing operators to zoom in for a closer look before pulling the trigger. According to the Army, this giant prop represents a revolutionary shift in how we will wage wars. This is the face of the robotic infantry. Unmanned ground vehicles (UGVs) have already flooded the battlefield. There are at least 6000 robots in use by the Army and Marine Corps in Iraq and Afghanistan. For years these small, remote-control vehicles have allowed troops to peek around corners and investigate suspected bombs. And while unmanned aerial vehicles have been loaded with missiles since 2001, the arming of ground robots is relatively uncharted territory. Last June the Army deployed the first-ever armed UGVs. Three SWORDS (Special Weapons Observation Remote Direct-Action System) robots landed in Iraq, each equipped with an M249 light machine gun. These UGVs are essentially guns on tracks, a variant of the remote-control Talon bots routinely blown up while investigating improvised explosive devices. When the trio was approved for combat duty, the potential for historic robot-versus-human carnage lit up the blogosphere. Never mind the dozens of air-to-ground Hellfire missiles that have already been launched by a squadron of armed Predator drones over the past seven years—this was a robot soldier, packing the same machine gun used by ground troops. The historic deployment ended with a whimper after the Army announced that the SWORDS would not be given the chance to see combat. According to a statement from Duane Gotvald, deputy project manager of the Robotic Systems Joint Project Office, which oversees robots used by the Army and Marines, “While there has been considerable interest in fielding the system, some technical issues still remain and SWORDS is not currently funded.” The robots never fired a shot, but Gotvald pointed out that the Army’s 3rd Infantry Division used them for surveillance and “peacekeeping/guard operations.” The nature of the robots’ “technical issues” remains an open question. The Army has not released details, and officials with Foster-Miller, the Massachusetts-based contractor that developed the SWORDS, refused interview requests for this story. But according to Col. Barry Shoop, deputy head of West Point’s electrical engineering and computer science department, the reason armed UGVs continue to lag behind UAVs is because of their mission: close-quarters firefights. “The technical challenges are greater,” Shoop says. “Think of the kind of image and graphics processing you need to make positive identification, to use lethal force. That’s inhibiting.”

AMMO IN ROBOT

THE SILVER BULLET:

It is shocking, but will it happen? The project has its critics, even in the Pentagon, where many doubt that technology can deliver such a "silver bullet". But the doubters are not in the ascendant, and it would be folly, against the background of the Iraq disaster and the hyper-militarised stance of the Bush administration, to write it off as a computer gamer's daydream.
One reason Washington finds it so attractive is that it fits closely with the ideologies of permanent war that underpin the "war on terror". What better in that war than an army of robot warriors, permanently cruising those parts of the globe deemed to be "supporting terrorism"? And what a boon if they destroy "targets" all on their own, with not a single US soldier at risk. Even more seductively, this could all take place out of sight of the capricious western media.
These technologies further blur the line between war and entertainment. Already, games featuring urban warfare in digitised Arab cities are everyday suburban entertainment - some are produced by the US forces themselves, while a firm called Kuma Reality offers games refreshed weekly to allow players to simulate participation in fighting in Iraq almost as it is happening in the real world.
Creepy as this is, it can be worse: those involved in real warfare may have difficulty remembering they are not playing games. "At the end of the work day," one Florida-based Predator operator reflected to USA Today in 2003, "you walk back into the rest of life in America." Will such people always remember that their "work day", lived among like-minded colleagues in front of screens, involves real death on the far side of the world? As if to strengthen the link with entertainment, one emerging military robot, the Dragon Runner, comes with a gamer's control panel. Greg Heines, who runs the project, confesses: "We modelled the controller after the Play Station 2 because that's what these 18-, 19-year-old marines have been playing with pretty much all of their lives."
The US aspiration to be able to kill without human involvement and with minimum risk raises some dreadful questions. Who will decide what data can be relied on to identify a "target"? Who will be accountable when there is an atrocity? And what does this say about western perceptions of the worth and rights of the people whose cities are no more than killing fields, and who themselves are mere "targets" to be detected, tracked and even killed by machines?
Finally, the whole process feeds alarmingly into the "homeland security" drive in the cities of the global north. The same companies and universities are supplying ideas to both, and the surveillance, tracking and targeting technologies involved are closely related. What we are seeing is a militarisation of urban life in both north and south that helps perpetuate the biggest and most dangerous myth of all, which is that technical and military solutions can somehow magic away resistance to George W Bush's geopolitical project