Efficiency. This single word encapsulates the objective of supply chain and manufacturing executives. The ideal state of manufacturing is to provide just-in-time (JIT), in-sequence, on-demand production. This ideal state-led Kiichiro Toyoda, the founder and second president of Toyota Motor Corporation, created the Toyota Production System (TPS) in the 1930s. TPS defined the guiding principles behind what has become known as lean manufacturing or simply ‘lean’. The overarching goal is perfection: eliminate all waste — or inefficiency — in the end-to-end manufacturing process.
Observing human behaviour, measuring manual processes and setting standardised work has been fundamental to manufacturing since the 1910s. Through these steps, supervisors and engineers gain deep insight into processes. But observation itself is extremely time-consuming, irredeemably biased and produces data on a very small scale. On the other hand, AI-powered observation and measurement are instant, impartial and create massive information datasets.
In addition to AI, other technology-driven advancements enabling digital lean include video traceability enhanced security and manual process analysis tools. These technologies all collect, analyse and regress significant amounts of data to reduce process waste. Before diving deeper into the critical digital lean technology solutions, it will be helpful to understand why digital lean is so important in the first place.
The case for digital lean
Because advances like AI continuously improve the process, digital lean offers manufacturers an opportunity to realise both speed and cost savings without sacrificing quality. When applied holistically, digital lean maximises overall process efficiency by improving all three legs of the manufacturing stool: quality, speed and cost, as discussed below –
Sacrificing quality for speed and cost improvement risks future business and your company’s reputation. Conversely, improving quality can elevate your brand image in the market. Toyota enjoys being the gold standard for quality in the automotive industry and embraces and trains other companies in its version of lean manufacturing, the Toyota Production System. Toyota has demonstrated that strict adherence to quality is critical to protecting your brand image.
Complete video traceability
For years, manufacturers have used object recognition technology and defect detection to ensure quality on the assembly line. Now, manufacturers have the option to take quality to the next step with complete video traceability. Suppose you have conducted design and process failure mode and effect analyses (FMEAs). In that case, video traceability provides vital insight into the validity of severity, occurrence and detection assumptions that populate your overall risk profile. This visibility reduces defects and scrap rates while increasing the critical efficiency that process engineers seek to optimise.
In-line quality inspection
Complete video traceability enables in-line quality checks with a critical reduction in risk and subsequent improvement in throughput. Digital lean connects inspection to production, eliminating the need to add another station to the process. This enhancement shortens the time to implement corrective action when needed and is more effective when paired with AI, automatically flag anomalies in a video and alert the line associate in real-time.
Less reliance on visual inspection
Though line associates are skilled at their craft, they are human. As a result, you have to plan for some amount of human error throughout the process. End-of-line visual inspection of the components is your first line of defence, but it doesn’t prevent defects from happening. Pairing inspection with video and data mitigates the risk of human error by providing redundancy in how you inspect the parts and preventing defects from making it to the end of the line.
With improvements in quality, getting these good parts made and shipped is next. Digital lean offers substantial benefits in essential operations like order fulfilment, inventory and demand planning. Adding AI enables automated optimisation of these enhancements, realising the benefit sooner than having engineers regress the data and modify the process.
Faster order fulfilment
Putting your product in the customer’s hands as soon as possible is the best way to delight them with your solution. Any time between receipt of the customer order and the start of manufacturing is process waste. Digital lean connects the ERP system to the production process, rapidly converting accepted orders to the beginning of production.
Automated JIT inventory management
There is nothing more disruptive than a force majeure at a customer site. Knowing the status of your inventory and ensuring your customer has sufficient (without excessive) supply carries a high value. JIT manufacture is one of the key facets to the ideal state of lean and implementing digital lean moves the process closer to that state.
Elastic production to meet elastic logistic demands
The supply chain was one of the hardest-hit disciplines from the coronavirus. Discretionary service-based functions came to a screeching halt while remote ordering exploded beyond even the most aggressive projections. Digital lean enabled adopting companies to react faster to this disruption than those who maintained traditional approaches.
Unconstrained demand for medical devices, such as ventilators and respiratory equipment, created dangerous shortages in items critical to treat folks who fell ill with COVID-19. Digital lean can collect essential data to alert manufacturers of a demand inflexion and rapidly respond. Many companies who experienced unconstrained demand pivoted to adapt to this elasticity; those who lagged in reacting lost market share to those who were nimbler. They could not match the growth rate of the competition.
Cost reduction through OEE & OPE improvement
With quality and speed optimised, profitability improvement through cost reduction is the third lever to achieve optimal financial efficiency with digital lean.
Machine Learning and simulation
AI and ML continuously improve the manufacturing process automatically. In the spirit of continuous improvement, AI and ML software ‘learns’ from the video and process data and produces analytics and insights accordingly. Traditionally, this feature accounts for machine and tool wear and improves the capacity of the capital over its useful life. Running the equipment at peak efficiency extends the life of the equipment further beyond its depreciation schedule, delaying cash outlays for new equipment and reducing the need for human intervention. All of these improvements maximise the ROI.
Simulation is another tool that can start the manufacturing process at its optimal point. Ideal for approach comparison and sensitivity analysis, simulating a run condition or capital fatigue can set the process toward its highest efficiency on day one. The massive data collected by digital lean can also help to tune the numerical models for improved predictability.
Optimised manufacturing flow and plant layout
The equipment’s arrangement can significantly affect both the OEE and OPE or overall people efficiency. Capital should be running near 100% of the time to maximise cost efficiency. Engineers can tie process changes to scheduled equipment downtime to improve OEE. They can design the plant layout to remove waste in manual processes such as hand-carrying a part to another station or to conduct a visual inspection. Digital lean automates these approaches and defines improvements to optimise both equipment utilisation and human and materials workflow.
Training for greater efficiencies
One of the most impactful benefits that digital lean provides companies is the opportunity for training. Arming management with quantified data and a visual representation of how manual assembly is happening illustrates the impact of recommended changes on the process. It’s a comprehensive digital twin of the factory that includes human actions, something previous digital twins have been forced to omit due to the lack of technology to gather this information.
The data shows them how to look at the process to optimise how they do things. One of the most extensive education opportunities is quantifying the benefit of JIT vs batch assembly. JIT is a pillar of lean, but line associates often gravitate to batch to work on one process step at a time.
Current digital lean technology driving AI-powered production
Digital lean provides numerous benefits in quality, speed & cost improvements, and engineers are beginning to extend the impact of current digital lean technologies by augmenting them with AI. The primary objective, especially for manually assembled components, is to minimise human error and adapt to process variability changes in real-time while matching throughput to demand. Below is a review of the leading-edge technologies and lean methodologies that are amped up with AI.
‘Go to the Genba’ is a core technique of lean: essentially, go to where production is happening to know what’s going on. While there is no substitute for laying eyes on the process, the pandemic stopped travel, creating the need for an alternative approach. Remote video recording and live streaming from the line provides a clear view of operations without the cost and time inefficiency of continuous improvement consultants travelling to various sites. Remote Genba is sure to remain popular even with travel back on the table.
Manual process analysis tools
Capital equipment can be connected to the central processor to increase performance and improve OEE, but companies need to implement additional manual process analysis tools to aid manual assembly. Implementing data-driven recommendations and backing them with a video account creates a clear, positive message to justify the process improvement.
Real-time and post-facto video traceability and search
Reducing defects & mitigating warranty and recall risks are the fundamental ways to reduce waste in lean manufacturing. For a quality engineer aiming to understand manual assembly problems on each assembled device quickly, video traceability can help you identify and isolate the cause of quality issues, speeding root causes analysis and ensuring only the impacted units are re-examined, not entire batches.
Cycle time data
Industrial engineers prioritise productivity, a tangible measure of efficiency. Suppose they can collect and analyse thousands of cycle times across their plant instead of dozens of data points that are typical today. Collecting cycle time data on this order of magnitude enables manufacturers to prioritise their human and machine resources better. Furthermore, they can eliminate the time spent on tedious and incomplete manual time and motion studies.
Process step data
Collecting manual process production data is fundamental to maximising overall system efficiency. Industrial engineers want to collect enough data to have an accurate picture of the number of process steps to make more effective decisions about standardised work and improve productivity. A step-change increase in the amount of data collected sharply improves manufacturers’ ability to predict and diagnose failures and impacts to efficiency. Employing video as part of standardised work also provides natural line associate training and learning new tasks.
Digital lean can assist quality management in validating the process for specified accuracy. It is nearly impossible for humans to catch every opportunity or perfectly adjust to each out-of-tolerance condition. The connected equipment and information on manual tasks can converge on process limits that meet the success criteria, saving time and money while increasing throughput.
The lean leverage on manufacturing assembly
Digital lean is the next iteration of traditional lean manufacturing and delivers substantial benefits. The data generated through the cycle, backed by video, provides manufacturers with a competitive advantage in the world of digital lean. Although manual assembly is woefully unaddressed by modern Industry 4.0 technologies, leading-edge technologies that improve lean methods, such as Drishti, are poised to revolutionise how manufacturers leverage manual assembly across industries.
Courtesy: Drishti Technologies, Inc