Robotics Case Studies
- This case study is filed under:
Pioneer Hi-Bred International Wins 1999 Robots & Vision User Recognition Award for Robotic Palletizing Operation
Pioneer Hi-Bred International is one of the largest seed producing companies in the world with headquarters in Johnston, Iowa, and research, production and sales organizations in more than 100 countries. Because Pioneer is an agricultural-based company, most of the production facilities are located in rural settings. Labor is hard to find and difficult to retain, especially with heavy, manual work jobs.
Pioneer has a number of product return 'rework' centers located within the central United States. These rework centers each receive many thousands of returns each year. Each bag is manually moved from receiving pallet, verified as to product type, and placed on the sorted pallet. The bags are then moved to storage warehouses. The stored pallets are returned to the rework area and removed from the pallet by hand, 50 pounds at a time.
All bags are palletized for original shipment to the customer. All return products arrive palletized. However, all depalletizing, sorting, inspection and restacking of the return product is currently handled by hand. This can occur three or four times depending on need. This means that most bags are restacked with the use of manual pallet stacking locations and manual labor. Pioneer does not want to require their employees to lift and manipulate many thousands of 50 to 60 pound packages each year.
The problem was identifying equipment capable of locating, picking and moving the flexible, partially filled craft paper bags used by Pioneer. The bags are made of three-ply craft paper with a poly-liner. The bags come in two lengths - 31' and 33'. The bags have varying gusset thickness - 2 ½', 3', 3 ½', 4' and 4 ½'. Each pallet contains eleven layers with six bags per layer. Four bags are positioned side by side, and two bags are positioned end to end across the end of the four bags. Each layer is a mirror image of the one above or below. At this point, tie layers are not used.
The bags are filled with flowable material of varying weights. The bags can weigh as little as 42 pounds or as much as 69 pounds. Since each bag has a sewn upper and lower lip, the bags are not flat, put actually tapered at the top and bottom. Side to side curvature is caused by the gussets -making vacuum pickup of rounded surfaces difficult.
Pioneer uses the same bag sizes for different products. Each bag has a different product/bag space ratio. Because of this, bags with similar gussets and lengths might actually have different thicknesses. One bag may be very full and rigid while another bag may be partially empty with a loose surface and unfilled corners and gussets. This complicates determining the location of the bag tops from layer to layer.
Since all bags are moved by hand, this means that most are restacked with the use of pallet stacking locations. The fixed walls of the pallet stacking areas help maintain a square pallet but they do not affect the skew of the bags as positioned upon each pallet layer. Each bag position may vary significantly from the adjoining bag.
Rate of Bag Movement:
Finally, the most difficult problem was to perform the work mechanically while maintaining or exceeding the rate provided by hand labor. This meant the rate over time must equal at least 20 bags per minute.
Pioneer worked with integrator Automated Concepts (Council Bluffs, Iowa), to find the least expensive and most applicable equipment for the task at hand. The equipment located and installed included conveyors, frames, tables, and other necessary parts and pieces.
By intent, Pioneer limited its search to only those parts and pieces currently in the marketplace. While Pioneer has the ability to design and develop equipment like it desired, it felt that everything needed was available in the marketplace, but simply not being used in the capacity required.
The integrator provided the programming required to make the robot function as requested. This programming, when combined with imaging equipment from DVT Corporation (Norcross, Ga.), made the final system work well and exceed original expectations.
The facility being affected:
Before, during and after the integrator began to help design the system, the facility being affected spent a great deal of time working on the parameters that needed to be met. This helped win some initial acceptance and provided some buffer during those times when things were not working as well as hoped. The minimum criterion for the project was a test cycle rate of 12 to 15 bags per minute. The local facility warehouse team established these criteria. This rate was achieved in a test cell that is far from being optimized for proper layout.
The improved robotic system:
The robot and the machine vision systems are an integral part of the process required to locate, pick up, and remove the craft paper bags from a 42' x 54' wood pallet. The vision system determines the location, angle and elevation of each group of three bags. The robot then moves the vacuum pickup end tooling into place, picks up the bags three at once, elevates and rotates the bags and places them (one at a time) upon a conveyor leading to a bag breaking operation.
Economics of the Installation
The system has functioned for one complete rebag/ rework cycle - approximately 6 months. To date, the system has functioned more than 90 shift days, and moved in excess of 450,000 bags of product.
Location # 1 - Test Cell
- Reduced first shift labor by 800 hours
- Reduced second shift labor by 600 hours
- Reduced per bag rework cost from 44 to 28 cents per unit
- Reduced injury, down time, errors
- With reduced labor requirements the facility was able to start the rework process sooner and move to JIT supply of product.
- Planned internal rate of return on this test system is over 18%
It must be pointed out that this depalletizing system is only the first phase of a three-phase plan. Phase II will prove the equipment can palletize. Phase III will prove the equipment can depalletize, sort by bar code identification and re-palletize each product in separate locations.
Each of these phases will use the same equipment and the same physical space. Each of the phases, when viewed separately, has an Internal Rate of Return greater than 18%.
Problems - Mechanical or Other
Accurate repair and maintenance records were maintained throughout the season. Total repair parts totaled less than $500 material and labor. Total 'down time' for repair and maintenance was less than 18 hours.
At the beginning, many employees were afraid of the change. Everyone was assured no jobs would be lost. None were lost or are anticipated to be lost. No employees want to return to moving 50-60 pound bags by hand. In fact, they resent the few bags that still require manual movement and are working to find a way to handle those.
There were no mechanical or electrical robot problems or vision system problems due to equipment failure.