Manufacturing + intelligence = magnificence


Manufacturing Intelligence Solutions

A production manager’s headache

As a production manager in manufacturing, there is increasing pressure from the executives higher up to increase efficiency, reduce costs and provide feedback on measurable outcomes such as CPUs and KPIs.

Whilst you may have sensors on your equipment and some data available, it’s a schlep to bring it all together, analyse it and generate reports that provide a comprehensive view of your facility in real-time. In fact, it’s a major distraction from doing your actual job: overseeing production processes, sourcing materials, coordinating operations, etc. Wouldn’t it be easier, and less headache for you, if a computer could do all that for you?

Well, the great news is that there are software systems that fulfil those requirements and a few that are capable of even more. But where do you start? Well, you’ll need big data storage and real-time streaming, then there’s the analytics side of things in a variety of forms and data visualisation to make that easier to conceptualise. All of which needs to be compatible with each other and your current system. But what about when you need to add in another data set, such as time and attendance software or a second facility? Do you need a system that can scale as you do? A production manager plans for contingencies but if you’re not a software expert and you don’t have time to research all of the alternatives how can you do that? Do you feel like the production line is running too fast all of a sudden and you’re the one creating the bottle neck?

How about a headache pill?

Our unified data analytics platform, Locstat LightWeaver®, provides everything you need for real-time, intelligent production line decision-making. It is data source agnostic, which means it can aggregate information from all of your current systems (plus any new additions) and pipeline it into a central big data repository for processing and analysing. So no more complications with “off the shelf products” that aren’t compatible with supplementary data sets.

An additional benefit is, as a South African company, you pay local (not American or European) prices with the added bonus of in-country support. As a quick to implement, light touch on your existing architecture, you can make the change to next-generation solutions with ease.

That’s the simple answer but for a technical mind, you need more information than that. I know I would. So here’s the long version.

What we do

Each manufacturing plant is unique. At Locstat, we take the time to understand your business so that we can programme the system to ask pertinent questions of your data. We start with an audit questionnaire to establish the ‘as is’ in your facility, which includes questions like:

How many manufacturing sites do you have and do you have oversight of all of them? Does each site manager have real-time, data-driven visibility of their production line? Are you set up for job, batch or flow production and can you monitor these? How many production lines do you have? Do they operate simultaneously? Do you have industrial sensors and software systems generating data along your production line? If so, what, how, where and why? Do you have multiple, potentially disparate data sets coming from these sensors – but need it integrated? Have you mapped out your production line from a data perspective? What are your main issues and concerns? Have you identified the root cause of any maintenance problems? Do you have early warning systems in place? Are you reactive or pro-active? What are your specific KPIs (goals)? What metrics do you use to assess these? How are you visualising and reporting all of this?

And we keep going until we establish the perfect solution for you.

By collaborating with your subject experts, we ensure that the Locstat LightWeaver® Manufacturing Intelligence Solution provides the insights that you need for your specific business requirements. This includes mapping contextual metrics for KPIs and determining rules for condition monitoring and predictive maintenance.

Locstat Process

The production line – a quick overview

The production line is a series of complex machines working sequentially as an integrated whole and synergistically with their human counterparts to produce quality results efficiently and consistently. At least, that’s what we hope. In reality, there’s a lot of variation on that ideal and plenty to go wrong.

Machinery: We all know the horrors of a breakdown at an inopportune moment. You’re late for an important meeting and the car won’t start; your computer crashes taking that essential, time consuming and overdue report with it; crucial equipment fails when you need to complete a large order for a new client in record time. Then there are potential interruptions for regular or unplanned maintenance – some machine parts might wear quicker than expected or take more stress on certain production batches than others. This can suspend processes at inconvenient, or even costly, times. The simple solution to allay these issues is well-known: condition monitoring and predictive maintenance.

And let’s not forget the hours spent on set ups or changeovers (clean up, set up, and start up), including the hidden factor of the operator’s psychological adjustment to assembling a different product. The benefits of quick changeovers are cited as reduced defect rates and inventory costs; an increase in production flexibility; and an improvement in on-time deliveries. So it is definitely something to strive towards for greater efficiency and profitability.

The human component: Each line worker has a specific set of tasks to complete within a given time frame. But as we all know, individual levels of productivity can change from day to day or even hour to hour. Sometimes, we feel super human and at others we can struggle just to get through the bare minimum. Apart from ensuring good work practices (e.g. regular breaks, proper training) and safety protocols we don’t really have much control over this aspect. Or do we?

Maybe you’ve realised that some employees like regularity and others prefer change. Generally, the former will work consistently; whereas the latter will thrive at the beginning of a shift, after a break or a changeover and slow down as boredom hits. The time it takes for each one to get bored differs from person to person and day to day. However, monitoring sections of the line and individual workstations can provide insight into how each staff member operates. Then you can create a more dynamic work environment for those who need it (thereby increasing their efficiency) and determine whether time allocated (usually based on industry timing standards) is sufficient for the given task. Perhaps regular bottlenecks are not due to operator inefficiency (or boredom) but insufficient training or heavier workloads at specific stages of assembly. Without this understanding, it’s difficult to resolve the issue.

Some traditional stumbling blocks

“The difference between a stumbling block and a stepping stone is how high you raise your foot.”

Benny Lewis

First the technical side of things

Whilst we have covered some problem areas, there is also a big data topic to discuss. IoT or PLC sensors and asset tracking (RFID) are fairly standard in the industry, however, an increasing issue for many manufacturers is not a lack of technology but an inadequate or legacy IT structure. These systems struggle to process and analyse the sheer volume of disparate generated data in appropriate time frames. If the data sets are not integrated into a larger perspective and elevated to provide essential insights to relevant parties, then intelligence is not available to inform decision-making at the propitious time.

Is that headache returning?

Now from the human perspective

Any change can seem overwhelming at first. For a creature of habit there’s the comfort of “if it ain’t broke why fix it?” But if you can’t even see what’s wrong, you’re operating in the dark. Perhaps you just can’t imagine what improvements would look like or simply don’t know where to start?

The main concerns of adopting new technology are compatibility with current architecture, expense, lack of expertise and lengthy implementation times. This can result in huge uncertainty, expensive false starts and valuable time wasted. However, we can alleviate these pain points for you.

A magnificent manufacturing solution

A complete real-time view of operations provides the capacity to make informed and intelligent decisions about your business as issues surface. Then decision makers at every level can take effective and pre-emptive steps to maximise production, reduce waste and minimise disruptions.

When we say complete, we mean everything: equipment sensor feedback (on processes and for condition monitoring), inventory status, scheduling, quality control, order status, etc. All of this data needs to be incorporated into a central hub and analysed for a holistic view of your ecosystem. Streaming analytics is perfectly suited for real-time feedback on this type of high volume, interconnected, fast data. Add in a graph database component with the capacity to drill down for granular problem solving and you have the ability to create action steps to resolve on-going issues.

Of course, we can’t forget KPIs. These are critical in keeping track of your production processes. Are you meeting them, exceeding them or falling behind? Assimilating these into your analytical dashboard provides an enhanced perspective for intelligent decision-making. Some KPIs might need to be broken down into smaller metrics to simplify root cause analysis and identify problem areas early. For example, Machine Downtime, Unscheduled Downtime and Machine Set Up Time KPIs all contribute to Overall Equipment Efficiency (OEE); and First Pass Yield, Rework and Scrap factor in to the Quality KPI. A strong computation engine along with complex event processing makes all of this a natural, yet necessary, part of your strategy and is one of the core strengths in our Locstat LightWeaver® Manufacturing Intelligence Solution.

Activity Based Costing (ABC)

And what about Activity Based Costing (ABC)? Whilst this concept has been around for decades, it is not always easy to implement without supporting and applicable data. With the increase in automation and accompanying reduction in direct labour, traditional costing systems are less relevant in determining specific operational costs. ABC allocates overheads to multiple cost pools, e.g. purchasing, storing, machining, which are determined by cost drivers, e.g. number of purchase orders, machine hours or employees. This gives a more accurate assessment of expenditure and loss, providing a clearer picture of where cost-related inefficiencies need to be addressed and insight into how to achieve this. To run an ABC costing output, though, the data sets from the various activity pools need to be aggregated and feature engineered as some cost drivers or activity pools may require different weightings to accurately reflect their significance to your business. Graph technology can handle this with ease.

However, it doesn’t help to see issues if you can’t do anything about them. The ability to adapt to circumstances as they happen, change production schedules and mitigate risk hangs on a responsive system that can learn from historic data, contextualise it and provide recommendations on possible outcomes. This is where Machine Learning (ML) and complex event processing shine. The human element is essential here both for programming and ongoing knowledge. Engineers can make edits in the process (e.g. to improve yield) and ML will learn from the resultant data and, where necessary, pass it on to the next user. This circumvents some of the issues with skills shortages and augments the knowledge of newer employees.

And wouldn’t it be magnificent to fearlessly embrace innovation? Digital twins of your machinery and your plant allow you to virtually create current and proposed products and processes to assess viability, resolve issues (e.g. scheduling, unplanned downtime) and improve operational efficiency. This provides dynamic process control so that you can change tack if and when necessary. An added benefit is the reduction in waste (you can try out a variety of designs before making a physical prototype) and a shorter time-to-market for new merchandise.

With all of this in one unified data analytics platform, Locstat can make your data-based headaches a thing of the past!

Make the change and revolutionise your production line.

Magnificent manufacturing with Locstat!




You might be interested in …