INDUSTRY SCENARIO – WHITE GOODS

Human-AI collaboration in quality control of White Goods

Quality control at the ZHT dedicated laboratory at the MWO Factory in Cassinetta, ltaly © Whirlpool

The COALA solution will be deployed and validated against three industrial use cases:  (1) white goods, (2) textile and (3) detergent productions.

The Whirlpool use case is focused on the implementation of the COALA Digital Intelligent Assistant (DIA) to support the activity in the Zero Hour Testing (ZHT) laboratory of the Microwave and Oven (MWO) factory.

Objectives

Adopt a predictive quality strategy

link the quality control of the finished product with design stage and shop floor

reconfigure its production line and facilitate root cause analysis

Expected Impacts and KPIs

Key benefits are driven by the possibility to adopt a predictive approach to quality strategy:

• 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.

• 5 Stars (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.

KPI today

• Number of calls per factory (average): 10,000/year

• Number of high level quality issue per factory (average): 2/year

• Number of service intervention at final customer (average): 5,000 /year

• Organizational costs (excluding rework):5.2 FTE = 250,000 €

Expected impact 

• Number of calls per factory (average): -30% (-3,000/year)

• Number of high level quality issue per factory (average): 0/year

• Number of service intervention at final customer (average) : -30% (-1,500/year)

• Organizational costs (excluding rework): -20% (-50,000 €/year)

Business Scenarios #1

AI-assisted Quality Control in ZHT Laboratory – ZHT OPERATOR

AS-IS situation

• ZHT Operators executes the quality control based on his own experience and memory, referring to a per-defined and statical quality checklist. At any new product introduction a dedicated training has to be setup.

• The information required during the SOP have to be retrieved manually from various information systems (data on product, specifications, defect descriptions,..) and defects are input manually

TO-BE situation

• The COALA solution will allow ZHT operators to be driven into checklist execution and supporting to easily retrieve the requested information about products and quality, leveraging on recommendation driven by predictive analytics

• Digital assistant will replace current manual digital input of identified defects into legacy systems, leveraging on vocal control

• COALA will support operator’s skills and competencies development, adapting  checklist level of details to the real information needs and helping learning phase from novice to expert

• The COALA solution will enable the possibility to capture operator’s knowledge and experience, making it available to Whirlpool quality organization to boost cross-functional improvement actions (production and product development)

Business Scenarios #2

AI-assisted Quality Control in ZHT Laboratory – ZHT LAB SUPERVISOR

AS-IS situation

• ZHT lab supervisor get access to ZHT operator activity through legacy systems data (DCS) and through the informal dialogues at the end of the shift. At the beginning of the shift he passes the key attention points to be addressed  during the working time (expected products to be controlled)

• The access to quality data in legacy system has to be done manually with high complexity in cross functional analysis

TO-BE situation

• The COALA solution will allow ZHT lab supervisor to strictly monitor quality control process and drive to higher effectiveness through the recommendations resulting by predictive quality approach (quality risk assessment and checklist management) . He will be able also to monitor the effectiveness of quality control suggestions and of the resulting improvement actions through the impact evaluation on quality KPIs 

• Lab supervisor will be facilitated in managing operators skills and competencies development with faster and more effective learning process. to be driven into checklist execution and supporting to easily retrieve the requested information about products and quality

• COALA solution will provide also an effective tool supporting root cause analysis, and the possibility to capture and capitalize the knowledge and expèertize of ZHT laboratory to support quality continuous improvement process

What are the main challenges on the quality check process of the finished products so far?

The ZHT is a statistical quality check on functional and aesthetical aspects applied in a dedicated laboratory environment out of production flow on some finished products (MWO).

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

This ZHT procedure is then executed in the mass production phase of the 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.

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.

How will COALA improve Whirlpool quality performance?

We expect that the COALA Digital Intelligent Assistant (DIA) effectively supports the ZHT operators to execute the quality check on finished products as well as to capture potential quality defects, to be identified by COALA’s predictive analytics function based on the available quality data set and afterwards to suggest the ZHT operators the right actions.

Currently, the ZHT operators execute the quality control based on their own experience and memory, referring to a predefined and statistical quality checklist. When a new product is introduced, a dedicated training has to be setup. For each product to be tested, COALA automatically executes a quality risk assessment on available historical data to provide recommended additional control steps in case of identified high quality risk for a specific defect.

The COALA DIA is expected to replace the current manual digital input of identified defects into the legacy systems or of positive control result by leveraging voice control. The operators can accept or refuse recommendations provided by COALA. In case of refusal, they should provide an explanation. Their feedback should then be used by COALA to improve future recommendations.

Moreover, COALA will support operator’s skills and competencies development by 1) adapting checklist level of details to the current knowledge level of the operator and 2) transferring knowledge of experts to novices.

The COALA solution will allow capturing the operator’s knowledge and experience, making it available to Whirlpool quality organization to boost cross-functional improvement actions (production and product development).