Automating data collection removes the potential for human error and improves the accuracy of your data, promoting better decision making. Adding automated alarms can increase response times to out-of-specification products and reduce the risk of recalls by stopping production faster.
As data is collected across products, lines, facilities, and suppliers, it must be unified in a centralized database. Through the power of cloud computing, plant floor data can easily be uploaded to a single repository, but make sure all plants have standardized naming conventions so the data can be aggregated.
Once the data is unified, it’s possible to visualize more than the quality of a single product, line, or facility. You achieve real-time visibility of the entire enterprise, from end to end—including suppliers, incoming inspection, raw materials, in-process checks from shop floor operators and the quality lab, process data, packaging, and finished products.
With this visibility, a quality intelligence solution can reveal insights that were previously blocked by silos or locked in filing cabinets. The actionable information, generated about the enterprise’s processes, suppliers, and manufacturing operations, can identify opportunities for improvement.
According to a study by Honeywell, “respondents said they believe data can enable well-informed decisions in real time (63 percent), limit waste (57 percent), and predict the risk of downtime (56 percent).” But unfortunately, in the more than 30 percent of respondents that don’t have predictive/data analytics in place, and don’t plan to do so in the next year, many believe their companies can grow without it, that they already have what they need, or the benefits are overrated.
Is $1 million in sustainable annual savings overrated?
Applying the operational insights consistently across the enterprise streamlines, optimizes, and transforms processes and operations. For one manufacturer, eliminating overfill on 12 production lines in a single facility saved more than $250,000 in two years. When applied to multiple facilities, the sustainable annual cost savings exceed $1 million.
Re-evaluate Strategies for Improvement to Multiply Savings
Specified weight requirements in the food industry are a key performance indicator and regulatory requirement. If packages are filled too much, the cost of ingredients increases from overfill and negatively impacts profitability. Alternately, under-filling incurs the wrath of regulatory organizations and consumers alike, destroying your brand’s reputation for “cheating” the customers.
For one food manufacturer with more than 200 products and a three-shift, 24-hour operation, in-spec weight variances didn’t present a problem. But within the quality data they collected, there was a massive opportunity to improve profitability. After a thorough analysis of quality data and operational processes for each of the company’s top 15 products at one plant, summaries of net weight data over a 12-month period were compiled to compare performance across all production lines.
Based on the data in the summaries, different “what-if” scenarios were proposed to determine the best opportunity for cost savings. In the first scenario, the company considered what would happen if they took a traditional approach of reducing fill weights by 5 percent and maintaining current processes. This would lead to a $160,000 annual cost savings in raw materials. However, this would likely put some of the packages underweight and is therefore not feasible.
Instead, what if the variability in weights was reduced across all production lines? Based on the visibility and insights attained from the collected data, it was clear that each product ran differently on each production line (see chart). This variation was wide-spread, from line-to-line, shift-to-shift, week-to-week, and month-to-month. While few products were actually out-of-spec, the negative impact of variability on the overall plant performance was distinct and typical for an organization that had never before visualized or summarized data across the enterprise.