お問い合わせ

AI and Problem Solving: The Ongoing Role of Humans in Manufacturing Operations

Each robot added in manufacturing replaced about 3.3 workers in the US according to a recently released study.Using data from 1990 to 2007, economists Daron Acemoglu of MIT and Pascual Restrepo of Boston University were able to quantify the effects of automation on the labor force. The trend is accelerating.  In his book, A World Without Work: Technology, Automation, and How We Should Respond, Oxford economist David Susskind argues that on the frontiers of artificial intelligence, machines are no longer replicating human capabilities. They now use vast amounts of processing power and data to solve problems in ways humans cannot.

Many basic “troubleshooting” activities can now be resolved without operator intervention

Indeed, machine learning has allowed organizations to make amazing strides in automated problem solving.  Many basic “troubleshooting” activities can now be resolved without operator intervention. If the system, the data and the corrective actions are all well-defined and accurately documented, it reduces the load on an organization’s resources in dealing with relatively standard issues. In turn, this allows an organization’s human capital to focus on complex issue resolution activities that still allude the machines.
Machine learning through the accumulation and analysis of vast amounts of data allows for quick, machine-driven solutions to many common problems. But, at least for now, it also means that problems still need human involvement and these problems usually are complex and unique. Despite, or perhaps because of, the vast advances of digital manufacturing, the need for advanced problem-solving skills required throughout manufacturing operations persists and is higher than ever before.

The need for advanced problem-solving skills required throughout manufacturing operations persists and is higher than ever before

For the present, the problems that commonly remain beyond machine learning’s ability require a high level of human problem-solving proficiency in certain areas. Machine learning algorithms can lack the ability to think beyond their intended application. When problems occur beyond an algorithm’s intended scope, they may require these types of specific, human responses:
  1. Reasoning Ability.  People can offer the ability to analyze why a particular situation is happening: the difference between common cause and special cause problems.  Why is this happening in this particular case?
  2. Contextual understanding.  People are needed to interpret various issues or human interactions, understand how they impact or cause a situation and determine how this context may play a role in resolution of a complex issue.
  3. Ability to identify new information. People with advanced problems-solving skills have the ability to identify specific pieces of missing information that are relevant to issue resolution and to develop and execute a plan for obtaining this information.
  4. Ability to detangle complex situations.  People can take a group of complex issues that are interrelated and read the situation, ask questions to clarify the issues, prioritize specific issues, and then build teams / assign responsibilities and manage the teams to resolution.
Advanced problem-solving capability has surfaced as a priority for operations personnel at all levels. Some of the basic capabilities needed for addressing complex problems within a manufacturing organization include:
Setting priorties. By narrowing focus down to the essentials that will solve a problem—information, skills, problem-solving process—accelerates the ability to address the immediate problem or the immediate customer needs.
Reducing problems to manageable parts. By effectively dividing the problem into parts and solving these individual parts before connecting them to make a whole reduces complexity.
Using a process. Working within a well understood and defined problem-solving framework increases objectivity and accelerates time to resolution.  This involves using a common approach, common “language” and the visible use of structured, problem-solving thinking.
Collecting data. Is the data needed available and where is it? Problem solvers need to use various collection methods, identify relevant data and sift through the noise of high volumes of data.
Building proficiency. Repeated practice and use of effective problem-solving processes help to build a high level of competency.  Proficiency is promoted by defining expectations (what does good look like) and comparing this to capabilities and achievements as they grow over time.

About Kepner-Tregoe

Software and templates don’t solve problems. People solve problems!
What kind of people? People who are curious, ask great questions, make decisions based on facts, and are empowered to lead. They remain focused under pressure and act confidently to do what needs to be done. You’ll find these problem solving leaders both at our clients and here at Kepner-Tregoe. For over 60 years, Kepner-Tregoe has empowered thousands of companies to solve millions of problems. If we can save millions for a manufacturer, restore IT service for a stock exchange, and help Apollo 13 get back from space, we can help your business achieve success.

ブログ画像1
Three Tips for Navigating the Ocean of Data to Improve Problem Solving in Manufacturing
ブログ画像1
How to solve problems when you are missing “required” data
ブログ画像1
The Future of Work – Data-driven decision making
ブログ画像1
The future of problem solving in manufacturing

お問い合わせ

お問い合わせ、ご意見、詳細確認はこちらから