By Tom Jensen
ELAU Inc., a Company of Schneider Electric
For many years two factors gave robot designers and manufacturers a lock on developing equipment for the packaging market: patents and the specialized kinematic knowledge required to program robotic motion. While the robotic arms were under patent, the controls held the unique motion algorithms needed to handle the complex path planning, blending, and resolution of multiple trajectories to the same point. Thus, robot articulating arms and specific controls were exclusive to robot developers.
The patent on the 3-axis delta robots has run out, enabling others to develop robotic installation kits based on this popular design.
Conditions have changed, however. The patent on the 3-axis delta robots – those recognizable ‘spider’ arms typically mounted over a workspace and used for pick-and-place with light objects – has expired. The 3-axis delta robots and the two-axis versions are used in both primary (product pick-and-place) and secondary (cartoning and case packing) packaging systems. Now, these arms are available as ‘kits’ for those who wish to outsource or design their own.
As for the controls, no longer are robotic control systems exclusive to robotic manufacturers. Open architecture systems not only replace robot controllers, they also replace the packaging machine’s controller and the PLC as well.
It is not “plug-n-play” simple yet to take these “do-if-yourself” kits and create your own robotic system. Therefore, here are a few tips to ease the process of creating your own robot-based
Application considerations. Articulated robotic systems traditionally suit case packing and palletizing, with smaller versions handling carton erection, filling, and sealing. Gantry and portal systems are typically used for traditional palletizing and for handling heavy payloads and lower speeds.
The first step to ensure that a robot offers the necessary freedom of movement required for these actions is to select the right arm for the job. The 3-axis delta handles high speeds and light loads, a category expected to increase as the concept entered the public domain in Europe a year ago and the U.S. in late 2007. The 2-axis delta handles heavier payloads, offers deeper reach into cases, and collating. It is a popular arm for the current trend of applying robotics in secondary packaging. However, both versions may require additional servo axes to perform their functions.
Motion control considerations. The advanced mathematical algorithms ensure smooth coordination of a robot arm’s multiple joints, wrist actions, and linear travel. While some simple gantries use point-to-point positioning, the real efficiencies come when motions are fluid, fast, and focused on the tool center point (TCP). It is this point that differentiates machine design and robotic motion design.
In robots, motions relate to the TCP rather than to individual axes. Motions are defined by the target position and type of movement of the robot arm through a set of trajectories for each servo axis. The trajectories are individually calculated and synchronized by a virtual camshaft. The various mechanical components operate in unison and can be adjusted dynamically.
Because robots can develop considerable G forces, a feature such as intelligent acceleration monitoring can limit accelerations and velocities and contain the resulting centrifugal forces.
The controller calculates the required trajectories for each motor at runtime, enabling the tool path to be easily changed at that time.
Different degrees of freedom are possible, depending on the control software and the manipulator within a given envelope or workspace. Cartesian movements are inherent to the control system. Both the motion and mechanics are flexible.
Commercially available development tools. Motion control toolkits have been introduced in the past three years to overcome the need for the specialized kinematic skills. The kits offer robotics libraries that you can use to program cartesian motion just as you would for a conventional machine into an IEC 61131-3 Function Block. The application of a transformation Function Block then performs all the necessary kinematics.
When all the machine functions are embedded in Function Blocks, it is possible to develop programs in a modular structure, which improves diagnostics, reusability, and response to inputs.
Robotic systems can then be designed as modules linked together through concatenated Function Blocks to perform the transformation (also know as “trafo”) necessary. For example, for a familiar 6-axis articulated robot, plus a trafo for the wrist movement and a trafo for the end-of-arm tool actuation.
As an ARC report recently stated, “Machine modularity allows machine builders to configure a packaging machine based on functional subsystems such as bottle carousels, labelers and wrappers. Integration of a robotic manipulator further leverages the concept of modularity…”
What to look for in development kits. Robots are capable of developing some G forces, and too much can overcome the gripper’s holding force on the product. Therefore, look for intelligent acceleration monitoring to limit accelerations and velocities and contain the resulting centrifugal forces.
The motion commands should include point-to-point, linear or circular interpolation, and splines. Spline curve algorithms map a continuous path between start and target points. Look for a geometric blending capability to reduce cycle times by ‘blending’ the path to optimize speed and distance traveled to reach the target point. You should be allowed to define your criteria for target point, velocity, acceleration, and jerk.
The program should keep both forward and backward movements on the same path – and this is harder than it might first appear, like backing up in your car at 60 mph. Likewise, it should stay on path during an E-stop.
The program should also be able to precisely trigger peripheral motions, such as indexers, wrappers and sealing mechanisms.
Other considerations. It is simpler and the response time faster to integrate robotics with packaging machine operations when control is centralized in a single controller. However, the controller must then be powerful enough to manage one or more robot arms plus all related functions, such as belt tracking and vision systems.
Many controls vendors today claim robotic control among their bag of tricks. But several features can vary greatly, including: ease of development, ease of operation, integration with the rest of the packaging machine, and the response time to adjust an ongoing motion, such as changing belt velocity and just plain speed. Thorough testing will illuminate the differences among suppliers.
ELAU Inc., a Company of Schneider Electric
Nuspark Engineering, Canada, Case packer
The compact case packer from Nuspark Engineering can be equipped with one (shown) or two side-by-side 2-axis delta robot arms depending on throughput requirements.
Infeed systems can be synchronized to transport, collate, orient, and buffer products, as well as provide positive control of case opening. Such benefits are possible from embedding robotics in machinery.
The additional arm has no impact on footprint. The design uses servo modules, and is therefore scalable without increasing the size of the electrical cabinet.
Fallas Automation Adabot R700, USA, Case packer
Fallas Automation calls its design the Adabot because it lets you easily add a ‘bot — actually up to four case packing robots together — all running off the original controller.
Fallas selected the 2-axis delta for its higher payload capacity and deeper reach into cases. Their design uses stiff, lightweight carbon fiber to maximize payload and speed, resulting in 80 cpm for loads up to 2 lb and 40 cpm for payloads to 5 lb.
Fallas specializes in difficult to handle bagged and flow wrapped products, and is configured to pick uncollated packs at these speeds without the use of a vision system.
Robotic flexibility, including wrist rotation, allows the Adabot to produce different pack patterns that would require two different conventional machines. The operator can build these patterns from the HMI and save them as recipes for reuse.
Cavanna Cartesio G35EFC, Italy, Cartoner
The Cavanna Cartesio Model G35EFC robotic cartoning system integrates three articulated robotic arms to form, fill, and seal cartons in one unit with a footprint of about 16 ft x 6 ft x 6 ft.
Each articulated robotic arm uses three servo motors. Seven other servos tightly synchronize an infeed belt, an oscillator device, a dual-belt racetrack collator, and a flighted carton infeed conveyor. All of these are synchronized not only with each other but also with the motions of the three robotic arms.
A traditional mechanical machine would not be able to deliver the smooth and precise positioning accuracy of this robotic machine.
Pester pac robot 3, Germany, Orienter/infeed system
Pester Pac Automation’s pac robot 3 is another 2-axis delta design. It’s shown here with their PEWO-pack 450 Compact bottle wrapper for the health and beauty care industry. Continuous product guidance allows gentle handling of up to 300 unstable bottles per minute.
As bottles are discharged from the labeler, they enter the collating unit where the pac robot 3 picks two sets of six bottles, indexes them 90° and places them on the wrapper’s infeed belt. Bottles are inserted into the servo-controlled, single-lane sealing unit at up to 50 cpm, and wrapped and sealed in PE film before entering the new PEWO-therm shrink tunnel.
Robot programming tips
While not conclusive, here are a few points to keep in mind when programming robots.
— A coordinate system is a way to describe the space around the robot arm. CRS robots, for example, can use any of four coordinate systems: World, Joint, Cylindrical, and Tool. Coordinates of any of these systems describe any point in robot space.
— The origin of the world coordinate system is the center of the base of the robot. In some cases, if the robot is to be hung suspended above the workspace, you may need to change this origin. To do so, set a base offset. A base offset is a set of coordinate values that redefine the new origin.
— A location is a specific point in space stored by the robot for use in an application. Types of locations are cartesian location (cloc) for coordinates based on the world coordinate system, and precision location (ploc) for coordinates based on the joint coordinate system.
— Straight-line motion is defined as multiple joint motions synchronized so that the tool axis moves along a straight line. Straight-line movement can initiate from any position along specified World or Tool coordinate systems, but is not available in the Joint and Cylindrical coordinates systems. You can also move in a straight line to a specified variable if it is a Cartesian location. You cannot move in straight-line mode to a precision location. Straight-line movements are useful for maintaining the level or orientation of a payload.
— Even though robot motion is controlled by servomotors, you can disengage servomotor control of the robot joints with the limp command. When limp, the robot joints can be moved by hand or by an external force. Individual joints can be limped, or all of the joints can be limped at once. When an axis is limp, the encoders still supply feedback to the controller. Caution is required with limp
motion. A limped joint is affected by gravity or inertia, which can result in robot collisions that can damage the robot or other equipment.
— When moving along an axis or to a specified location, the robot moves the tool center point (TCP). You can change the TCP with a tool transform command. A tool transform informs the controller of the position of the tool center point (TCP). Without a tool transform, the controller moves the arm as if the TCP is the center point of the tool flange surface. A tool transform consists of the measurements in the tool coordinate system of the mounted tool’s TCP. The tool transform also includes the yaw, pitch, and roll coordinates that define the tool’s orientation.