The problem-solving process: A modern, data-driven approach for operations managers
Global competition requires organisations to rethink and continuously improve their processes, including the problem-solving process. In the manufacturing operations environment, problems such as equipment failure, product defects or minor stoppages may occur, and teams need the knowledge, skills and confidence to tackle these problems quickly and effectively to prevent recurrence and minimise disruption.
Data is one of the most important assets for manufacturing operations right now, and emerging technology such as high-performance software tools help both employees and machines to harvest and process the data essential for effective problem-solving. This data-driven approach speeds up the problem-solving process, and can help implement long-lasting improvements that add to the bottom line.
Follow these steps for a more data-driven approach to problem-solving.
Key steps in a data-driven problem-solving process
1. Assemble the team around a digital dashboard
Employees such as the machine operator, maintenance person, team leader and quality technician will typically make up the problem-solving team. They should meet preferably in the production area where the problem occurred, and gather around a digital board for quick reference to real-time data. Visual management is a fundamental part of problem-solving, as well as team empowerment and ownership. When displayed on a digital board, data can be analysed in real-time enabling teams to make more informed decisions.
2. Define the right problem
The team defines the problem using, for example, a technique known as 5W1H (five ‘w’ questions – what, where, when, who, which - and one ‘how big’ question). Make sure you’re solving the right problem. Toyota is justifiably famous for applying its problem-solving ability to perfecting its production methods. According to Toyota, the key to its method is to spend more time defining the problem and less time working on the solution.
The definition of the problem is the foundation for all of the work that follows. If you don’t clearly define the problem, you may go off in the wrong direction and waste time and resources as a result. Make sure the problem statement contains a clear articulation of the gap between the current state and the end goal.
3. Extract and analyse the data
The ability to analyse vast amounts of structured and unstructured data to gain insights underpins most digital transformation efforts.
Problem-solvers need the time, skills and training – such as analytical reasoning, data literacy, data science and data engineering – as technical and methodological knowledge is essential for extracting and analysing the data.
The challenge, then, is to extract the right data. Automation such as lineside systems can assist with extracting machine or production line data (loss data, for example). These systems identify specific issues and proof of a problem, as well as the data needed for effective root cause analysis. Once resolved, it provides evidence of improvement and could even draw attention to opportunities for further improvement.
The lineside system feeds live data via the cloud, which allows the problem-solving team to perform real-time analysis using graphs and charts displayed on the digital dashboard.
4. Find the root causes
Root causes may be related to people, machine, method, material or measurement, and using a data-driven approach to root cause analysis often provides new insights – in addition to the benefits already discussed. It also enables a more holistic approach where successes can be implemented throughout the organisation, rather than providing isolated solutions.
With the results of the data analysis, the team starts brainstorming the root causes. There are numerous methods to determine root causes, such as the cause-and-effect analysis – also known as the fishbone diagram. The 5 Why technique, however, is frequently used since it is simple and robust enough for use by all levels in the organisation. By identifying the problem and then asking ‘why’ five times (sometimes more, and other times less) – getting progressively deeper into the problem – the root cause can be strategically identified and tackled.
These Lean techniques are still relevant, but they are now supported and their time of application is accelerated by a data-driven approach.
DMAIC (Define, Measure, Analyse, Improve, Control) is a data-driven problem-solving methodology often used in manufacturing to reduce process variation and eliminate defects for long-term improvements. Integrating predictive analytics – attributed in one survey as the technology with the single greatest manufacturing impact – across the DMAIC process provides good insight into how each improvement phase is working.
5. Document, decide and implement the solution
The team documents the root causes and possible solutions. They could take these solutions to the next daily operations review (DOR) meeting for discussion, and there it gets decided whether or not to implement the solution.
Once the solution has been decided, it has to be implemented by your team. So keep it simple. You should be able to explain the solution clearly in 30 seconds. Limit the action items to solve the problem to about three. Think 80/100: Go for the solution that solves 80% of the problem, but that is 100% implementable by the team — rather than the 100% solution which is unlikely ever to be properly implemented.
6. Solve less complex problems first
In situations where you have multiple problems to solve (for example, during a value chain transformation which is ideally supported by an integrative improvement approach), pick the low-hanging fruit first. Solve the less complex problems. This approach gives momentum, shows progress and gives your teams confidence.
7. Let your team solve its own problems
As an operations manager, you need a strategy for how a problem is approached and managed. Anticipate the unexpected and use the strengths of your people to ensure that the strategy leads to a sustainable solution.
In the organisational design of the future, the hierarchical structure of centralised decision-making and problem-solving — where, for example, only managers, engineers and other specialists are the problem-solvers — is replaced with work teams that apply a much more collaborative, dynamic and agile approach to problem-solving. Without this new approach, problem-solving is more difficult because you will deal with individual contributors rather than team players nurtured in a cross-functional environment. Decentralised problem-solving frees up some of your time to focus on other activities that will drive the organisation to success.
A problem-solving culture encourages work teams to be proactive rather than reactive, and solve problems as they arise. As an operations manager, continuously ask questions such as “What is the problem you’re trying to solve? What’s actually happening? How do you know that? Why is it happening?” The aim is to find solutions and make the company stronger.
At first, you’ll need to do a lot of coaching: Create a safe environment where anyone can suggest a solution without criticism; develop problem-solving and analytical skills within the team; encourage questions to find the root causes; build self-esteem and team spirit; and ask individuals to evaluate their level of participation in problem-solving – which will motivate them to increase their contribution.
Data transparency and accessibility will help build confidence as individual contributions are backed by data displayed on the digital board.
Problem-solving is a great catalyst of growth and opportunity within the organisation. Although existing approaches help to solve problems to some extent, an advanced, data-driven approach helps identify root causes faster and increases certainty, while potentially offering a significant impact on overall business performance.