INDUSTRY SCENARIOS – WHITE GOODS PRODUCTION

Scenario 1: Accurate End-of-Line Quality Control with Human-AI Collaboration

This scenario will address the end-of-line quality control with human-AI collaboration with the aim to adopt a predictive quality strategy that will link the quality control of the finished product with the design stage and the shop floor. By integrating all available information sources (e.g. sensor data, historical operational data, and expert knowledge), it will be able to a) predict low-quality products and processes (e.g. at which workstation the defect starts occurring) and b) to plan mitigating actions. In this way, Whirlpool will be able to reconfigure its production line to proactively mitigate the root causes of the defects in order to, among others, reduce organization, warranty but also reputation costs.

Baseline. In the Whirlpool production model, the whole white goods production is tested from quality and safety point of view in order to ensure a high standard level of product quality to final customers.

These type of tests (100%) are executed in all EMEA factories either through the usage of automatic dedicated machines at the end of production line (It’s the case of functional and safety testing) or through automatic, semiautomatic or manual checks in some critical workstation along the production flow (visual quality checks, quality gates). 

To these testing actions, Whirlpool Production system adds also some statistical quality check actions that are applied both on internal production parts on quality critical processes (Statistical process control stations) and on finished goods, after the packaging process. In particular way this last testing, called  Zero Hour Testing (ZHT) is referring to the Statistical Quality Control applied in dedicated laboratory out of production flow on some finished products (generally from 1-3% of each production order) retrieved from the quantities ready to be delivered to the markets.

It’s main objectives are:

– Measuring the quality level of the outgoing product from an aesthetic, functional, and normative point;
– Measuring the effectiveness of process control.

These test are executed in dedicated laboratory environment, created in each production site, and following a specific STD operating procedure defined during Product Design phase. Factory Quality Managers are fully accountable of the procedure execution while Central Quality department is accountable for the procedure designing with Product Engineering department.
This testing method is designed to replicate the customer approach to product, simulating the normal product usage conditions at final customer first usage that is following the instructions in the usage and maintenance booklet / Instruction For Use (IFU), executing the suggested product installation and testing the first usage in home conditions. This method includes also the key elements of craftsmanship, considering the product appealing, the visual and touching characteristics, the noise, and also the conformity with the defined BOM (Bill Of material):

– aesthetic rating and static functional rating; 
– dynamic functional rating; 
– conformity with the Bill Of Materials rating. 

Currently, the procedure is executed manually by a laboratory operator and it is fixed, statically defined during process design phase both for what concerns checklist and reference parameters, and for statistical product withdrawal rate. 

These type of tests (100%) are executed in all EMEA This ZHT procedure is then executed on mass production phase of production lifecycle as a prescriptive list of elements that is rarely modified, generally only in case of relevant quality issues triggered by KPI monitoring and control alerts or by relevant market issues. In any case this review is, generally, not re-looping the product design process, limiting the reaction scope on mitigation actions at production level and reducing the possibility to anticipate similar quality issues for other products in development.

In the same time the quality issue facing procedure is activating only after the issue discovering at market level with impact on customers appreciation, losing the possibility to anticipate the issue for customers and to properly manage the risk. 
This approach, not guaranteeing the identification of all the defected products exiting from production lines, leads also to products that experience No-Fault-Found (NFF) effects in the warranty period. This late identification during usage leads to increasing costs of warranty policies and customer dissatisfaction. Moreover, the quality control at the very last stage of the production process constitutes mainly a reactive approach aiming at averting the delivery of defected products to customers and at correcting inefficiencies in the production process for the next batches.

Using COALA Solution

The COALA solution aims at adopting a predictive quality strategy that will link the quality control of the finished product with the design stage and the shop floor. By integrating all available information sources (e.g. sensor data, historical operational data, and expert knowledge), it will be able to a) predict low-quality products and processes (e.g. at which workstation the defect starts occurring) and b) to plan mitigating actions

In this way, WHR will be able to reconfigure its production line to proactively mitigate the root causes of the defects. 
A second important contribution driven by acronym solution is in the facilitation of root cause analysis applied on existing data availability that, due to the high complexity of white goods production process (high number of components, several quality critical process within production flow, high production pace, high variety of product range), is very often very difficult to be executed at shop floor level. An interactive system collaborating with human operator and leading into a deep dive analysis, will facilitate the identification of early signal of quality derailing effects and the possibility to consolidate a knowledge to be used for the future.
An augmented human-AI interaction through a DIA is of outmost importance to enhance the shop floor workers’ capabilities in identifying potential failures, investigating root-causes, and addressing the causes effectively, allowing them the opportunity to fully leverage on whole data availability to anticipate the events instead of reacting the events.

We expect this will optimize the overall manufacturing process and the product design process towards a zero-defect strategy. These effects will be measured by KPIs, such as 

·       GSIR (Global Service Incident Rate) = number of service calls in the first year of sales.
·       1MIS (First Month In Service) = number of calls in the first month of warranty.
·       PEX (Product Exchange) = number of products replaced to customer within warranty period.
·       5Stars (Customer product appreciation)
·       QLoss (Quality Defect rate at shop floor)
·       1MIR (First Month in Reliability) = number of defects identified in first month in short reliability test.

NFF problems are widely common for products with a long operational life. It creates costs attributed to the event investigation, testability of the suspected components, and maintaining a component support service [14]. When a quality issue arises after factory production release, either in the markets logistic network or, even worse, by final customers, a complex information management and issue root cause analysis starts, involving several functions:

Consumer service
·       Receives and manage the claim from customers and activate service partners to address the issue /repair, explanation, product exchange (cost per call, pex, spare parts & labour).
·       Monitors and controls calls data to verify if there are high frequency on specific topics, signal of potential epidemic (0.1 FTE for each market – 35 markets).
·       In case, activates alert process upstream to central quality and to markets logistic.

Market logistics: on alert evidence manages product blocks for orders processing and supply management (0.2 FTE for each event).

Central quality: activates quality verification on other markets and initiates focused projects in Manufacturing to address quality issues (0.5 FTE for each event).

Manufacturing: activates at central and local (factory) level focused projects to manage root cause analysis and fix the issue (1 FTE for each event).