It looks like one of those brain teasers, where a jumble of rectangles of varying sizes have to be assembled into a cube. That's how the Liquor Control Board of Ontario approached the challenge of automating a key aspect of its distribution-center operations.
The target was the supposedly mundane task of palletization. It’s an essential step in getting product out the door, but it can be costly and labor-intensive – especially when the cases being handled are of multiple dimensions.
High volumes only complicate the challenge. An Ontario government enterprise, LBCO is one of the world’s largest buyers and retailers of alcoholic beverages. Its biggest D.C., the Durham Retail Service Centre, handles some 50 million cases per year – 60 percent of its total volume. On-site inventory amounts to around 2.7 million cases annually, involving more than 8,000 SKUs that are destined for 653 stores throughout the province.
The fulfillment system is basic enough. Store orders are grouped into batches, broken down into pick orders, and assigned to pick stations. Conveyors automatically move each case to the appropriate store-order accumulation lane. Cases of multiple sizes arrive at the lane in random order. Workers then manually construct a stable pallet, building up the tiers in the manner of a 3D puzzle. The task requires 68 employees in two shifts, managing approximately 250 cases per man-hour.
At least that’s how things used to be done at the Durham RSC. Given a lack of automated solutions for palletizing cases of mixed dimensions, LCBO had no choice but to rely on the heavily labor-intensive method. Yet it was determined to cut costs at the facility, which was 30 years old and designed to handle half the case volumes it’s saddled with today.
High labor expense wasn’t the only issue. The Durham RSC was experiencing a large number of worker compensation claims due to back-strain injuries. Moreover, the stability of the pallets varied widely, according to the experience, skill level and degree of care exercised by employees at the assembly locations.
Blending Old With New
The challenge was to devise a solution that would automate the palletization of mixed-sized cases using traditional, tier-oriented pallet loaders and conveyance systems. “We didn’t want to reinvent the wheel,” says George J. Soleas, executive vice president of LCBO.
It was vital from the start to engage all stakeholders in the effort, he says. In addition to forming a project core team, LCBO reached out to three external vendors: Millar Systems Integration for software, Norpak for hardware design and implementation, and Columbia Machine Inc. for the palletizer equipment.
Also in the loop were LCBO’s retail outlets and, perhaps most crucially, the Ontario Public Service Employees Union. LCBO needed to get the union on board for a project that would on its face result in far fewer workers handling palletization duties.
The automated system would consist of six distinct components that address both the physical equipment and the software needed to run it. Key to the latter was the development of a sophisticated algorithm that could account for all characteristics of the cases that were flowing through the Durham facility.
The formula would need to create theoretical patterns, then emulate them and evaluate their stability in the real world. It would group cases by size, calculate optimal sub-patterns, then fit them into blocks to form the tiers of the finished pallet to a predetermined height. In all, the algorithm passes through eight independent phases, beginning with “request tier pattern” and culminating in “select best tier pattern.”
On the hardware side, engineers needed to incorporate such elements as an upstream device to measure the dimensions of each arriving case, as well as a buffering area for temporary storage of mixed-size cases waiting to be formed into tier patterns.
According to Soleas, the project required two man-years – two people working for 14 months – to develop the algorithm and conduct the feasibility study. That was followed by the construction of two prototypes, and rollout of the equipment to select store-order accumulation lanes.
Keeping It Running
Getting the new system up and running in a live environment wasn’t easy. For the Durham RSC, it was business as usual during testing and installation. And the optimization algorithm had to be integrated into the real world. Soleas likens the challenge to “doing open-heart surgery on a runner while he’s running a marathon.”
In the end, 15 manually staffed palletization lanes were integrated with eight machines, performing work that was previously distributed across 28 lanes. Each of the machines can palletize 1,250 cases per hour, five times faster than the old manual process. In addition, says Soleas, the system has resulted in the building of denser and more stable pallets, less product breakage (by an estimated 60 percent, with all lanes in operation), improved traceability, a safer work environment and better service to LCBO retail customers.
Some 77 percent of the facility’s total volume is now being handled by the automated palletization machines. Headcount, as expected, is down sharply, from 68 workers in two shifts to just 10 in the 15 automated lanes.
With all eight machines up and running, the Durham RSC has been able to eliminate 75,000 direct labor hours – an annual savings of $2.2m a year, says Soleas. By reaching out to the union at the outset, LCBO avoided potential labor strife, even as it dramatically cut back on staffing.
“We’ve not laid off a single person because of this project,” Soleas says, adding that staffing cutbacks have been achieved entirely through attrition and the elimination of casuals.
Soleas claims LCBO was the first to combine an automated palletizing system for mixed-size cases – an innovation that could easily be extended to other types of products that move in pallets. For its part, LCBO is now rolling out the technology to its warehouse in London, Ontario.
Liquor Control Board of Ontario
Timely, incisive articles delivered directly to your inbox.