Does Pricey Robotic Construction Make Good Sense?

Thu July 29, 2004 - National Edition
Pete Sigmund

Could a single person, from a remote spot such as a home, operate a whole fleet of construction equipment?

Sure. The technology is here — the Global Positioning System (GPS) positioning blades of graders, stringless pavers following terrain exactly as a 3D model dictates, on-board microcomputers sensing obstacles, video cameras showing everything which is happening, and wireless remote control.

An operator in a control center, using radio, wireless, GPS, video equipment, and other technology, can direct equipment relatively easily.

Experts said costs of totally remote operation would be too high at present to make it worthwhile for widespread use, but that the technology is very viable and is already being used for specialized tasks dangerous for operators. They also predict an exciting, very productive, future, including more automatic operation, for computer-aided equipment.

Robots are already building walls of houses and researchers said they will someday take over every housebuilding task, from building walls to installing plumbing.

Other robotic equipment is being developed that may allow cranes to automatically place beams for buildings with great precision and stability.

Here’s what some of the nation’s leading experts in automation told Construction Equipment Guide (CEG) about the future of automated equipment.

Technology Marches On

Interviewed at their laboratories by phone, the experts agreed that the technology for totally automated construction, without operators on site, is available, now.

“The technology of remote operation is very well-known,” said Yvan Beliveau, head of the Department of Building Construction at Virginia Institute of Technology in Blacksburg, VA. “You can do all the controls remotely. You can replicate a cab, in what we call virtual immersion, so that an operator would think he was really out there except that he doesn’t get jostled around. Everything is very viable and totally feasible technically. It’s only human factors that have to be resolved.”

Computers would be at both ends. In a “multiple packet exchange,” they would process and decipher signals from the equipment and re-send control signals.

“The equipment senses its environment through strain gages and accelerometers, sends video information on what the environment looks like back to the remote operator, who then runs the machine as if he were there,” said Beliveau.

The operator could move the machine remotely by high-speed radio links, observing results on wireless video. He or she knows exactly where the equipment is through the GPS.

“If you were digging yesterday, you can go back to that same spot,” Beliveau said. “Or, say, you want to load a truck, you can have it follow a GPS trail back on its route.”

Alan Lytle, a construction robotics expert at the Intelligent Systems Division of the National Institute of Standards (NIST) in Gaithersburg, MD, explained remote operation this way:

“Instead of pushing a button in a cab, you can stand next to the equipment and make it move. It’s not that much of a leap to do it remotely at greater distances. It’s like moving a Tonka truck [toy] by cable or wireless instead of pushing it by hand. However, cameras and sensors in the equipment still must show you what’s happening, and the machine must usually be augmented with its own devices like collision avoidance systems.”

Some Large-Scale Applications

How, and when, would this technology be used in construction?

“I think parts of this have large-scale application,” Beliveau replied. “Full remote operation, however, is a little further away. Certainly tracking for trucks, to make sure they won’t harm anything, is feasible, using fail-safe systems so that they don’t hit anything. They wouldn’t necessarily be driverless, but they will stop, or not do things, if you do something stupid, like exceeding load requirements. The truck’s microcomputer would say, in effect: ’OK, if you put 250 tons on my 175-ton truck, I’m not going.’ All of that could be done through sensors. It would be very simple and a good way to protect the equipment as well. This type of automation is being used now and will become much more pronounced.”

Good for Hostile Environments

Beliveau believes that remote operators will be used principally in very specialized applications.

“We have lots of things now that help equipment operators be more efficient, but taking operators out of the cab and employing them remotely will be done mostly because of hostile environments, safety issues, or lack of trained personnel,” he said. “Remote operation is certainly useful for environmental cleanup, nuclear cleanup and contaminated environmentally hazardous sites. The technology is also being employed in a lot of mining operations.

“Pattern-recognition by on-board computers is now also being used with fail-safe systems to ensure the machines don’t run off cliffs or hit each other. Remotely or even self-operated vehicles are certainly happening, but not on the scale that I think people might have predicted a few years ago.

“Automation of the machine itself, however, will absolutely become more dominant. To be able to analyze equipment motion, see if it’s efficient whether you have enough equipment in the right mix — all of this will be done automatically.”

Cost Question

Virginia Tech’s Lytle told CEG that costs of totally remote operations are the big question.

“Sure, it [full remote operation] is technically feasible,” he said, “but what would be the advantage, the return on investment, for a construction firm to invest in that sort of expenditure? Crane operators are expensive so are you going to have everyone sitting at home and dialing into a construction site?

“It’s technologically feasible to do limited construction operations remotely, and, with research, you could expand that capability. What hasn’t been proven is the cost benefit. In environments where you have to pull humans out of the environment, however, it [total remote operation] is worthwhile.”

Lytle cautioned, “Before I take the human completely out of the loop, are there things that I can do to improve operation while he’s sitting there? A perfect example is machine control of graders and pavers. It has been deemed cost-effective to invest in the technology to assist the human in performing fairly precise operations. Does that mean you want to do that remotely? Do you want to take the human out of the grader? As you see a cost benefit, do you continue to provide automation capability to that equipment? When you reach a point where it’s highly automated, you can start analyzing whether it would be feasible and cost-effective to pull the human out.”

The RoboCrane

Back in 1999 and 2000, an NIST research project demonstrated the ability to remotely operate a six-degree-of-freedom crane through an Internet-type connection with video feedback and with sensors feeding a computer-graphics virtual environment.

“The operator could sort of fly around the crane payload virtually and see where everything was,” Lytle said. “The idea was for an operator to dock a steel beam into a holder between two columns using Atlas connectors. They demonstrated they could do that. Then they demonstrated that an operator could do this from another building, without being in proximity to this simulated work environment.”

Lytle and his associates are now experimenting further with docking steel beams autonomously (no human on the equipment).

“I tell the crane to go from one point to another; I press a button and it does it,” he explained. “A modification of that type of crane could be used in the future. We’re using a crane, which we call RoboCrane, which NIST designed to stabilize crane loads. Instead of just a single hook, the crane can sway, pitch and yaw around while it holds the piece steady.”

The RoboCrane is a highly-maneuverable gantry-type robotic platform, controlled by six cables, which can manipulate large loads.

Could something like this be used to operate a crane remotely? “Sure,” Lytle said, “but I don’t know the best configuration for this application.”

Is there potential for widespread use? “Not in its current configuration, but it can be modified,” he replied. “Is there value-added in implementing the type of crane that we’re using? We don’t know the answer to that yet.”

Lytle said the value-added could be in the stabilization or path control.

“I control the position of the platform by playing out, or reeling in, cables so it’s actually very stable,” he said. “The Air Force is looking at it for stabilizing containers during ship-to-ship transfers of cargo. Something like our experimental equipment can be retrofitted on cranes so there’s potential for improved stability in handling material whether you do it remotely or not. It’s technically feasible to operate the crane remotely, but there’s no proven cost benefit as yet.”

Peering Into Future

Smart equipment is already a recognized part of construction. Through on-board computers and software, excavators know what optimum loads they can carry, and can follow a pre-set path with the precise required movement. Concrete pavers automatically follow terrain exactly as a 3D digital terrain model dictates. Graders use the GPS of satellites to position their blades at exact heights within millimeter tolerances. Smart cranes compute and display lifting data almost like a spacecraft cockpit.

This is only the beginning.

Lytle believes that as more and more of technology pricing comes down, construction will see a lot more automation-related techniques.

He gives Laser Detection and Ranging (LADAR) as an example of advances in technology entering wider use.

“LADAR imaging of building construction sites is showing more and more promise and is actually being used quite a bit, much more than five years ago.”

LADAR systems, which have become much faster and accurate, create three-dimensional images of areas and objects. NIST is urging vigorous efforts to create a next-generation system, the size of a coffee cup, providing millimeter accuracy.

But the scientist sees a pretty wide-open field where anything can happen, commenting.

“Am I going to get robots flying around first? Is that going to be the primary thrust? I’m not sure. I think one of the areas where there’s a big push is ’What is the status of my site? Where are my parts, my pieces of equipment? How do I talk to them? How do I track them so that I don’t waste time trying to find them?’ We are going to see a lot of improvements in logistical tracking.”