How a Global Pet Food Manufacturer is using Ferry
September 4th, 2024
One global manufacturer of animal feed & pet food spoke to us about how they are leveraging data-driven applications to solve operational bottlenecks in their Canadian packaging lines, particularly around downtime & yield.
Challenges in Production
The site focused on packaging pet food had the following challenges:
Real-time attainment: production progress was tailored on pen & paper, and manually inputted into systems to record order completion. The site found it impossible to track production against the schedule in real-time to make adjustments and needed corrections. This resulted in a weekly production shortfall of 20%.
Downtime: stop reasons were infrequently recorded and not analyzed to identify improvements. This led to a production loss of $30k+ per month per line.
Micro-stops: there was no visibility into machine performance and how it varied across product types. There was a feeling of lost production opportunity cost but no way to quantify it.
Pen & paper processes: all production information was recorded on paper sheets, leading to transcription errors & difficult in analyzing information.
Yield: through manual infrequent checks, the site team had an indication of significant yield loss due to overfilling of bags. However, there was a lack of real-time information to help guide where yield losses were most prevalent across products, and when losses were occurring so it can be remediated in-process.
Putting the Operators at the Core
The company decided that it was critical to ensure that any manufacturing intelligence system must be designed with the operator experience at the core. Frontline operators are key to the successful & fluid running of production, and the site wanted to ensure that any new system was easy to learn, intuitive to use, and informative enough to provide real-time insights without requiring technical skills.
The site also wanted to ensure that a solution could be tailored to their particular processes and standard operating procedures (SOPs) on an ongoing basis. A dynamic, flexible solution was required.
The third key tenet of a required solution was to help reduce manual steps for operators when recording information, which were often error-prone. Production tracking and batch assignment were routine operations that were conducted by pen & paper, and not tied back into the broader reporting framework for the site.
Tying this all together, it was critical that any solution needed to be easily integrated into live operations, without the need for capex or significant time & resource investment. The packaging lines had some automation equipment but no PLCs, so the solution would need to be able to connect to legacy systems to identify production data. This would need to be tied into a framework where lots could be automatically identified, with operators being able to assign them to specific products.
New approaches to production management
The site team with colleagues in operational excellence are looking at a range of manufacturing KPIs & initiatives to improve operational efficiencies on their packaging lines:
Real-time production & batch tracking: live throughput monitoring with automatic batch recording for frontline teams to track against the production schedule.
Operator guidance: providing real-time feedback to frontline teams on how production is performing to help inform changes in schedule or process. Advanced recommendations on set-point setting for machines in response to identified yield or production losses can be utilized to adjust operations in-process.
Analysis-led management reporting: strategic insights into SKU performance, future capacity planning and shift allocation, within a broader move towards data-led decision making.
Data-driven continuous improvement: providing a set of tools for data benchmarking within a site that can applied across other facilities through operational excellence initiatives.
Outcomes
Within one quarter, the client identified core sources of operational inefficiencies that when combined would lead to over 30% increase in production.
Micro-stops and variations in line speed if increased by 1 bag per minute would result in a 10% increase in throughput, with potential upside of an additional 30% if target run speeds could be reached. The top ten most troublesome SKUs were identified as a source of productivity gains.
Downtime reasons were monitored & quantified, leading to the identification of a 15% increase in uptime by reducing changeover times and unplanned interruptions to production.
Finally, real-time yield monitoring resulted in the identification of a 2%-3% average yield loss due to overfilling of packages, which could be offset by better regulation of process setup.