Introduction to Digital Lean Manufacturing
Lean has played a significant role for the past few decades in driving efficiency across manufacturing organizations by eliminating waste and focusing on customer value. Lean techniques have promoted simple intuitive visual and analytical approaches to decision making, problem-solving and continuous improvement.
However, today’s advanced manufacturing technology and complex product landscape is creating new challenges for manufacturers that are sometimes beyond the simple manual methods. Thankfully, digital technologies have also become more practical for manufacturers and can be used to implement Lean principles in this complex landscape.
Smart Manufacturing and Industry 4.0 technologies including IIoT, cloud computing, AI and robotics make it practical to implement and integrate digital solutions for new levels of insight, accuracy, speed, cohesion, automation, and flexibility.
Digital alone is not a magic bullet. Nor is it a replacement for traditional lean. The key to driving down production costs while increasing performance is to find the right combination of digital and traditional lean—one that fits the organization’s current situation and desired end state.
The term “Digital Lean” is often used to identify this new era of Lean practices that stays true to the Lean principles while embracing real-time data-centric techniques from Smart Manufacturing for enhanced decision-making, problem-solving and continuous improvement. Lean can become part of the culture and mindset for an organization that aims to create more value for customers while reducing effort and resource waste in production activities. A culture that remains open to changing processes and continuous improvement of productivity and quality.
Industry analysts Deloitte, Bain and BCG have studied manufacturers over recent years and documented that the cost reduction among manufacturers who implemented digital technologies along with Lean techniques was 30-40% higher than among those strictly using traditional Lean techniques. [1, 2, 3] Below is a list of examples of how digital Smart Manufacturing technologies are being used to enable those higher levels of productivity and cost reduction.
Improved Decision-Making
Data Collection
Collecting more data in a timely manner is the backbone of Digital Lean techniques. The integration of OT (control systems, industrial networks, and systems) and IT (enterprise information systems) in Smart Manufacturing brings timely plant and production insights to the entire team managing and supporting production operations.
Data is collected in real-time via sensors, gateways, and smart platforms that make data available to production systems, analytics and AI-based algorithms that turn data into insights for better decision-making and trigger automated alerts to the right personnel to handle non-routine situations. Data is more transparent and accessible, making it easier to display, share, and analyze among multiple stakeholders.
New sensors, data gateways and software not only make it practical to collect data directly from machines, but also make it possible for machines to automatically identify products as they show up at the workstation. For example, using auto-ID technologies like RFID, a machine can load the appropriate program and tools without manual intervention. This type of automated changeover makes it easier to increase the product mix without affecting efficiency and allows the operators to focus on more value-added activities.
Manufacturing systems of record are often maintained via manual methods yet many value stream analysis forget to include the time wasted on inefficient paperwork and wait time introduced by paper-based problem handling procedures. Much wasted time and effort can be eliminated by applying digital techniques to these record keeping processes.
Metrics and Key Performance Indicators
An important step for the success of Digital Lean initiatives is to establish an agreement on Key Performance Indicators (KPIs) for the organization that encourage Lean behaviors. Standardizing metrics used in the organization helps to align the team on common goals and understanding of business priorities. Common KPIs for Lean include on-time delivery rate, cycle time, cycle time efficiency, asset downtime, cost of poor quality, and surplus inventory.
In traditional Lean, data for these KPIs would be collected over time, then later compiled, and manually analyzed in spreadsheets. With smart technology, this data can be collected automatically and continuously, allowing KPIs to be monitored in real-time or near real-time. This allows workplaces to quickly implement changes and adjustments to improve KPIs, as well as spot problems and waste early as it starts to happen.
For measurement of KPIs to be effective, the data must be correct, relevant, and consistent. Ensuring this usually means standardizing processes, making sure that the information coming in via smart technology is the most relevant, and that employees are correctly using the smart technology. Metrics work best if employees buy-in, trust the data, and actively participate at all levels of an organization.
Visualization and Display Tools
Once an organization has defined its metrics and collected data to measure them, the next step is to make the information visible and accessible to the multiple stakeholders in the enterprise including management and the team at the shopfloor. It is important for all stakeholders to understand how production performance impacts the customer and the business goals for the entire enterprise.
Traditionally data is gathered manually and recorded in spreadsheets for after-the-fact analysis. With smart technologies, dashboards are configured to automatically use collected data to generate charts, diagrams, and other displays that support everyday decision-making processes.
Old school Andon boards are replaced by digital Andon boards with dashboards updated in real-time at monitors mounted at hallways around the plant, as well as accessed through mobile devices and computers at any business office that needs access to the data.
Improved Resource Productivity
Resource Optimization
Simulation techniques can be used for what-if analysis to make decisions about what actions to take to optimize the use of resources and help achieve cost, quality, and schedule targets.
Centerlining is an approach to reduce process variability and increase machine efficiency in manufacturing. Manual centerlining processes include standardizing setup and quality practices, such as regular measurements by operators or inline measurements. With digital tools, operational data is collected directly from sensors or mobile devices and the corresponding machine or process adjustment actions can be performed quicker. AI-based techniques can automatically derive adjustments using historical data.
Optimization techniques that leverage real-time data include the use of artificial intelligence (AI) and machine learning (ML) algorithms to detect unusual patterns and trends that might be causing the use of more effort, resources and energy than usually required to perform the task.
Reduced Equipment Downtime
Unplanned downtime can cause big delays and time wasted waiting while equipment is repaired, and production is rescheduled to work around equipment issues.
Preventive maintenance is usually performed on machines based on a fixed frequency schedule based on maintenance manual recommendations. Advanced sensors for conditions such as vibration, force and temperature can be installed on machines, monitored in real-time, and analyzed by AI-based predictive maintenance algorithms to trigger maintenance only when needed based on machine usage and performance instead of a fixed schedule. The increased accuracy of methods using real-time machine data reduces the amount of downtime and maintenance cost.
Smart machines with embedded sensors can monitor their own internal conditions and are able to immediately alert workers of breakdowns, tool wear, or malfunction.
Improved Quality
Work Standardization and Worker Guidance
When workers understand the standard, they know the best way to do the specific job and avoid the waste that would result from deviations and errors.
Smart Manufacturing techniques strive to have the right data available to the worker at the right time (when it is needed) and in the right form to minimize manual search and reformatting processes. Data that includes the work instructions for complex production tasks and any data collected from machines that helps the operator make the right decision about next steps or adjustments needed to the process.
Virtual Reality (VR) techniques are becoming more popular to train workers on complex tasks without risking the impact of learner mistakes on the actual product, especially when the potential mistakes would be on high-cost items.
Digital work instructions are easier to maintain up to date than paper instructions. They can be turned into interactive tools to guide the worker as they perform the production job. Employees equipped with mobile devices, like smart glasses, can receive step-by-step instructions with illustrations or Augmented Reality (AR) visuals superimposed over the physical view of the production process.
Head-up displays that provide work guidance can also provide visual warnings when AI-based assistance software detects any anomaly in the process. These tools provide additional confidence for workers to take on more responsibility for self-inspection.
AI-based assistance can not only warn about quality concerns but can also warn the worker about safety concerns when handling hazardous materials or detecting hazardous conditions such as proximity to certain equipment like moving cranes or vehicles.
Reducing Errors and Process Variation
Digital Lean techniques can help with the quality challenge presented by the increased speed of change and complexity in today’s products. The cost of poor quality and reduction of defects are common types of waste tackled by lean techniques.
One of the techniques to increase quality is to add more in-process checks through either worker self-inspection or automated process monitoring with techniques like computer vision and AI that triggers warnings in real-time when anomalies are detected. This type of process monitoring, and warning system is equivalent to a digital poka-yoke technique.
Software solutions can track inspection and test samples to reduce the effort related to checks performed during the production process. For example, a smart screwdriver that captures torques values can be connected real-time to the production system, to not only let the operator know when to stop at the correct torque value, but also to automatically collect the data as part of the inspection records for the product.
When problems do happen, the data collected through smart systems allows for faster and better root-cause analysis. Mobile devices can be used by employees to submit error reports, observations, or suggestions for improvement from the shop floor.
The Lean technique of Jidoka refers to the ability to stop a process when an issue is found and immediately alert the supervisor and support personnel that can assist in solving the problem. Old school Andon lights used to indicate that the line was stopped are replaced by event triggered alerts that are displayed at dashboards and broadcasted via email to the respective support department to assist in clearing the issue as soon as possible.
Software solutions can help expedite the response to quality issues as soon as they are detected to quickly identify the root-cause and minimize the impact of the problem. The greater level of detail of data collected by smart systems makes it easier to find the root causes of problems and reduce waste by analyzing the relationships and patterns found in historical data. AI-based techniques can also be used to help identify potential root causes of failures.
Improved Flow
Production Flow, Line Balancing and Scheduling
The idea of maintaining a consistent flow in the plant and production process is one of the important principles in Lean manufacturing. The concept behind the lean technique of Heijunka is to not only level production effort throughout the plant but also to set the production pace to match customer demand patterns and reduce manufacturing waste by leveling the type and quantity of production output.
When the company makes products in smaller batches and maintains smaller inventory buffers there is greater flexibility to meet changing customer demand patterns. Digital production systems make it easier to track smaller product batches, change production schedules to match changing customer demand, and make schedule adjustments to work around issues causing unplanned bottlenecks and affecting the rhythm of operations.
Traditionally, the size of production runs for Heijunka are not recalculated often due to the complicated nature of calculating the ideal run size for each Heijunka wheel iteration. The use of data from integrated production systems makes it easier to recalculate the run size more often to match customer demand and account for current resource availability, production time, and quality expectations. This is especially true in a high mix production environment.
Demand Pull and Just-In-Time Inventory
Lean manufacturing encourages switching from planning and receiving materials and parts based on schedules to just-in-time methods that establish smaller buffers and signal demand pull based on actual inventory consumption. The manufacturer is purchasing materials based on actual versus forecasted needs. However, manual ways to signal demand can cause delays and errors.
Digital techniques that monitor inventory levels and automatically signal demand for replenishment based on actual inventory levels and lead times can improve accuracy and avoid delays while also maintaining low inventory levels.
Kanban is a visual cue mechanism used to signal, as material is pulled from inventory, that the stock level has reached the replenishment level and an order needs to be triggered. Digital Kanban can be implemented and automated or operator-driven triggers from the line can immediately send digital requests for replenishment of materials. Auto-ID technology such as RFID sensors can be used to track the consumption of parts in real-time and trigger replenishment automatically. Smart systems can also analyze usage trends to recommend changes to the replenishment levels based on actual parts usage and scheduled product orders.
A “control tower” system can collect data and direct material movement inside and outside the factory and integrated value chain. The production plans for each day can be adjusted by an AI-based scheduling algorithm based on the latest information on current factory conditions, orders, customer priorities, capacity, inventories, and supplier lead times.
Digital lean techniques can go further and enable efficient make-to-order processes to avoid the building of goods that sometimes end up on the shelf and not being required by any customer.
Kaizen or Hoshin Kanri?
We have discussed examples of how Smart Manufacturing technologies are helping manufacturers take Lean to a new level—Digital Lean. The next natural question is: where do we start? There are some general guidelines like start small, focus on return on investment (ROI), and look for “low hanging fruit” to achieve early successes. We can apply Kaizen processes to start quick with an aggressive view of continuous process improvement.
However, there is something additional to consider with Digital Lean initiatives. It is also worthwhile to consider a more strategic Hoshin Kanri approach to deciding priorities. The organization will eventually want to scale implementation of digital process improvements and integrate the new data sources across the enterprise. This is where a Smart Manufacturing strategy for the organization comes into play. Kaizen thinking alone can drive a culture of solving problems in isolation and could end up creating new silos of information if there is no concerted effort to create a platform to share information in the enterprise.
Many companies have started digital initiatives and stalled because they skipped the important step of linking the Smart Manufacturing initiatives to the business strategy. Hoshin Kanri is a method for ensuring that a company's strategic goals drive progress and action at every level within that company. Business goals provide the framework that holds the initiatives together. Continuous-improvement tactics are the muscles to achieve the business goals. Both are important.
If you do value stream mapping as part of your Kaizen or Hoshin Kanri analysis, make sure to include all the steps related to managing the paperwork and reporting that goes along with production operations. You might be surprised at how much time is wasted in the value stream due to the inefficiency of paper-based procedures.
Additional Resources
CESMII, the US Smart Manufacturing Institute, has established a set of tools to help manufacturers jump start their Digital Lean and Smart Manufacturing journey through alignment of business strategy and technology strategy for increased operational performance, customer service, and competitive edge. For more information on CESMII and the strategic roadmap building tools, you can visit: https://www.cesmii.org/education/roadmap-tools/
References:
[1] When Lean Meets Industry 4.0 - The Next Level of Operational Excellence, D. Küpper, A. Heidemann, J. Ströhle, C. Knizek, D. Spindelndreier, BCG, 2017
https://www.bcg.com/publications/2017/lean-meets-industry-4.0
[2] Digital Lean – A Guide to Manufacturing Excellence, P. Serlenga, I. Leppavuori, I. Moraes, M. Forlini, Bain & Company, 2019
[3] Digital lean manufacturing - Industry 4.0 technologies transform lean processes to advance the enterprise, S. Laaper, B. Kiefe, Deloitte, 2020
https://www2.deloitte.com/us/en/insights/focus/industry-4-0/digital-lean-manufacturing.html