Traditionally, automated guided vehicles (AGVs) have lacked flexibility, which has limited their ability to quickly and easily change paths or routes. But, new technology will allow the devices to make smarter judgments and more intelligent decisions on the plant floor.

Most driverless vehicles used for material handling applications are currently guided by laser technology, which has been around since the late 1980s. “It provides the best combination of accuracy, flexibility and reliability at a reasonable cost,” claims Mark Longacre, marketing manager and JBT Corp. (Chalfont, PA), which recently changed its name from FMC Technologies. “And, laser bumpers allow safe operation without contacting any obstacles.”

However, laser navigation has several shortcomings in production environments. For instance, there can be problems with narrow aisles and false reflections. Laser systems are also more expensive than other forms of navigation, such as tape or wire guidepaths. One possible solution to those challenges is machine vision technology.

“Today’s automated vehicles rely more on laser guidance, providing greater accuracy, but with higher installation and ongoing maintenance costs,” says Greg Cronin, executive vice president of Seegrid Corp. (Pittsburgh). “[Our] vision-guided robots are the first to offer the reliability, accuracy and flexibility required in an automated vehicle, but at a significantly lower cost.”

Seegrid has developed a new class of driverless vehicles called industrial mobile robots. They feature AGV-like capabilities with software and vision systems that can be “taught” to follow paths and navigate independently.

“This new approach represents a major breakthrough,” claims Cronin. “The technology relies on intelligent software and inexpensive onboard cameras to control the vehicle’s movement path. [Our] technology relies much more on internal software ‘smarts’ than the more proprietary hardware-centric approach of even laser-guided AGVs.”

A Seegrid GT3 tugger is currently used to deliver fasteners to the assembly line at a Freightliner truck plant in Gastonia, NC. The autonomous technology allows the device to be more versatile than traditional forms of driverless vehicles.

The Seegrid robot literally see its environment and learns from it. As a result, it can react much more quickly to changing needs or real-time activities on the plant floor. For instance, Freightliner engineers can easily change the parts distribution route from shift to shift. That allows the AGV to make 15 stops during the first shift, but only five stops during the second shift.

“The robots feature walk-through-then-work technology, which means they can be ‘taught’ by people taking them for a walk, instructing the robot along a desired path, adding behaviors, such as sounding horns and waiting at stop stations, and doing this training in a very short period of time,” Cronin points out. “[Our] robots use inexpensive modularized components that are easily field-swappable by local service technicians. [That means] service and support is much faster and affordable.”

While today’s AGVs have the ability to adapt somewhat to changing conditions, manufacturers are looking for machines that are even more flexible and more autonomous. The industry is moving toward vehicles equipped with much more sophisticated navigation skills, speed and precision.

“AGVs now have the ability to ‘learn’ their environment and adapt to changes within their environment,” says John Gallinagh, product manager for hardware products at Kiva Systems Inc. (Woburn, MA). The company recently unveiled a driverless system for warehouse and distribution center applications. However, it’s developing new features specifically for use on assembly lines and other plant floor environments.

“The Kiva Mobile Fulfillment System is unique in that it implements a bottleneck-free, parallel processing workflow, whereas traditional AGVs operate serially within an operation,” Gallinagh points out. “For example, traditional AGVs perform one task at a time and do not take on new missions until the assigned task has been completed.

“The Kiva system is fully adaptive and can be assigned multiple tasks that can be prioritized and reprioritized,” claims Gallinagh. “In addition, Kiva’s robotic drive units work as a team. If one drive unit is unable to perform a task, another drive unit will pick up the mission to ensure the timely completion of the mission.”

A new technology that’s starting to emerge is intelligent AGV systems (i-AGVs). They incorporate cutting-edge navigation technologies that, unlike traditional AGVs, allow true vehicle autonomy without any external guidepath networks or navigational markers, such as wires, mirrors or magnetic tape, to assist in their movement around a facility. They are not confined to specific travel paths that have to be taught or programmed into the vehicle.

“Intelligent AGVs maneuver about a facility using open-path navigation, a technique that uses features in the working environment as a frame of reference, ensuring that any pathway to a desired destination is available for use,” says Pierre Pinet, product manager at RMT Robotics Ltd. (Grimsby, ON). “There is no predetermined routing from point A to point B; the vehicle is only given the destination when dispatched.”

Based on its knowledge of the environment, which is embedded in onboard memory, an i-AGV independently decides the most efficient route to reach its goal, adapting en route to any obstacles it may encounter along the way, whether they are permanent, such as walls or fixed equipment, or temporary, such as boxes left in an aisle or workers walking through the area.

Pinet says i-AGVs are typically used in fleets “to provide functionality that can be best described as a virtual conveyor. They allow items to be transported from any location to any location in a facility, but without the fixed infrastructure normally associated with traditional material handling systems."

Pinet’s company markets an i-AGV product called ADAM, which stands for Autonomous Delivery And Manipulation. Its currently used in the tire manufacturing industry to transport work-in-process. According to Pinet, the system has already been adopted by three of the largest manufacturers in the industry, but he says there is also broad interest in the more general material handling industry.

Traditionally, AGVs travel at a slow pace for safety reasons. But, advances in vision technology will soon allow the devices to operate much faster. “Most AGVs travel 2 to 3 miles per hour, which can be a waste of time,” says Cronin. “By using new software and new sensing technology, our next generation of machines will go 7 to 8 miles per hour.”

Some AGV manufacturers are also experimenting with fuel cell-powered vehicles. For instance, JBT recently unveiled a vehicle equipped with an onboard methanol-based fuel cell battery charging system. It charges the AGV’s battery while the vehicle is in operation and when it’s parked. According to Longacre, 5 gallons of methanol can power an AGV for 24 hours. In addition, JBT has equipped an AGV with a hydrogen-based fuel cell that replaces the vehicle’s traditional lead-acid battery.

The next big technical hurdle will be to attach assembly robots to AGVs so that driverless vehicles can do more than just carry parts or work in process. “That’s something on the drawing board, but it’s still a few years out,” says Cronin.

Jim Caie, vice president of ARC Advisory Group (Dedham, MA), says onboard robotics technology has already been developed by engineers at General Motors Corp. (Detroit). The application uses two tabletop robots to attach small components in car doors.

“That makes a lot of sense for doors, because it’s a very complex subassembly,” says Caie. “The technology is feasible and it could be in use within two or three years, if business conditions improve in the auto industry.”

Robots would be used to assemble a door while it’s being transported to the final assembly line. Components such as wire harnesses, door locks, speakers and window regulators would be positioned and fastened while the AGV is moving.