Preactor helps Armstrong Floors keep planning and scheduling challenges underfoot.
Armstrong Teesside is one of Armstrong World Industries manufacturing units, a subsidiary of Armstrong Holdings, and is a global leader in the design and manufacture of floors, ceilings and cabinets. Armstrong Teesside has been on its present site for 30 years and is responsible for the manufacture of cushioned vinyl flooring products for the home, including the famous Rhino brand. The company produces approximately 18 mil sq metres of flooring per year, some of which is distributed within the group in the US with the remainder being supplied globally to flooring distributors. When Armstrong needed a production planning and scheduling system that would better enable it meet its corporate vision of “we deliver on our promises”, it found the perfect fit with Preactor.
Armstrong manufactures an 80/20 split of Make to Stock (MTS) and custom orders with the former having lead times as short as 5 days and the latter up to 7 weeks and being dependent on varying factors including a rotating print cycle and the number of SKUs involved. The company prides itself on its reputation for quality of product and also quality of service and applies the same commitment to delivering on its promises irrespective of whether an order is for 1 roll or 1 container comprising approximately 150 rolls. Given the continuous nature of the manufacturing process which runs 24x7 on a 5 day basis and very considerable set up/changeover times, meticulous production planning and scheduling is essential.
While manufacturing only involves 3 key stages, there are a sizeable number of permutations at each as well as a high degree of sequence dependency. First the substrate is created from fibreglass and the specific combination of saturate and gels required by the end product. This substrate can be in 1 of 3 different sizes – 2 for the EU, 1 for the US – with a finished rolled being approximately 1500 linear metres. This then passes to the print line which is where the complexity significantly increases as Armstrong has over 300 designs to choose from, each of which can be available in a variety of colours. With each design involving up to 6 sequence dependent rollers weighing in excess of 1.5 tonnes and requiring changeover/cleaning/setup times of approximately 30 mins, batching becomes extremely important. To complicate things further, because the substrate also comes in 3 different widths and several formulations, this adds a further setup/changeover consideration. From this stage, the printed and embossed product moves to the coating process where 1 of 17 different coatings can be applied, involving up to five sub-processes before the finished product is QA inspected, cut to length, packed, and stored in Armstrong’s 3.5 million sq metre warehouse prior to dispatch.
Perhaps surprisingly, the most pressing business challenge that Silvergate faces is the sheer pace of business that the company has to deal with, processing 1000 orders per month, all of which can have entirely variable delivery dates. This is exacerbated by the MTO nature of the business that means the company carries no finished goods stock. Visibility is critical; especially the means to accurately know how any job is progressing at any one time. As Bestall remarks, “Nothing’s static, everything’s always changing, and keeping track of what’s happening, where and when can be a full time job.”
Jan Longhorn is Armstrong’s Supply Chain Manager and has been with the company for 6 years. She describes the not inconsiderable planning and scheduling challenges the company faces. “The biggest factors we have to contend with are the number of SKUs - currently 2500 – and the large variance in order sizes we have to contend with. Given the continuous nature and sheer physical scale of the actual manufacturing, there is a large potential for bottlenecks. Even with elements of buffer capacity built into the machines to cope with short term stoppages or problems, each process is very sensitive to changeover and setup times, with even a minor adjustment in the plan potentially creating a sizeable bottleneck or simply moving the bottleneck from one place to another.” To put this into context, the print line in Armstrong runs at most for 55% of the time with the remaining time being taken up with changeovers.
It was Bestall’s appointment as Business Manager that brought about the beginnings of change at Silvergate. “We had to do something because we couldn’t adopt the manufacturing and sales strategy that we wanted without changing how we did our planning and scheduling.” The first step on this evolutionary journey was a drive to improve internal business processes by first mapping them accurately. This led to Bestall identifying 30 individual steps to process an order and a total of up to 60 steps from receipt of an order through to the order being delivered. This equated to 8 hours of time and represented a major point of waste within the company.
Other challenges arise from the nature and supply of the raw materials. Plasticizers for example are generally in short supply, with availability and cost being very sensitive to price fluctuations in oil. At times only a certain supply may be allocated by the producer, irrespective of Armstrong’s order book. Some raw materials are ordered on a 3 month lead time basis which can create problems if Armstrong receives an unexpectedly large order and has to deliver in a much shorter timescale. Clearly balancing forecast and stock levels are important but Longhorn elaborates on further challenges the company faces in this area. “Given the scale of manufacturing, we simply don’t have space for an excessive amount of raw materials or Work in Process (WIP). We therefore have to run fairly lean and as some raw materials can be delivered twice per day day, it only takes one late delivery to potentially stop the whole plant.”
Another reason why the company has to keep WIP to a minimum is that a range of inks used in the process have extremely short process lives and must be used quickly. If kept too long as WIP, the entire batch can be contaminated which means the entire batch has to be scrapped. Armstrong can’t compensate by producing an excess of a certain product because while the product itself has a very long life, in terms of sales shelf life, any given product may only remain in fashion for 1 year, after which it is generally un-sellable. As Longhorn notes, “Add in planned and un-planned maintenance, tears in the roll mid process, which if occurring during the heating phase requires 6 hours of cooling before the problem can even be begun to be rectified, and you have a fair idea of the challenges we face.”
As with many manufacturers, Armstrong did the best with the planning resources it had prior to investing with Preactor which meant Longhorn having to co-ordinate sales, raw materials, 3 different production lines, warehousing and dispatch. In addition to this, maintenance both planned and unplanned, new product developments and factory tests also had to be taken into account. Longhorn even had to be aware of shift pattern possibilities and how the availability of overtime might be affected by local factors such as a home football match. She puts her ability to balance all these things down to a mixture of “Excel spreadsheets, gut instinct, and internal company knowledge” although also admits to, “many times wishing I had a crystal ball on my desk.”
Whilst looking quite a manual process, Nolan is quick to point out that Silvergate wasn’t looking to fully automate the system at this stage. “We deliberately set about using Preactor as a means of evolution, not revolution. We wanted to enable our planners to do their job more efficiently and effectively. We also knew that we had a large cultural change to effect, both in terms of getting people to trust a piece of software as opposed to their own experience or gut feeling. This shift is essential because people often plan according to what they want to make, whereas we needed to plan purely according to our customer requirements.”
On a practical basis, planning was done on a weekly basis with Longhorn spending Tuesday through to Friday checking inventory levels and looking at the previous 13 week’s sales orders to establish what needed to be made. A weekly meeting was held during which any other influencing factors on the plan were discussed, such as maintenance, delivery problems etc. To physically construct a plan from all this information, Longhorn had created a range of complex and non-interlinked spreadsheets, each of which had a different general set of permutations which she manually had to manipulate and compare just to find out if a given production plan would actually work. As she readily admits, “Any changes meant a complete manual regeneration of the entire schedule and a repetition of all the comparison processes. Because everything was disconnected, we had no visibility of the impact of one factor on anything else. With no ‘what if?’ capacity, we couldn’t accurately work out the best response, we literally just had to work out any response and stick to it and then deal with the consequences if we got it wrong.” The location of all this expert knowledge within one person not only created an area of potential weakness for Armstrong, it also restricted Longhorn’s own career development as she was too valuable to move on to develop her role in the company.
The benefits were noticeable right away, most noticeably in the area of increased visibility. Silvergate could now see all the orders in the system in real time, and how each was progressing. When a new order was received, the impact of this could be seen, and Preactor could re-order the schedule right away as required. The time savings were also significant as the company moved from a service level of 79% for delivering in full and on time with a 5-7 day lead time, to 96-99% in full and on time, with no lead time. This demonstrable improvement also helped encourage a greater sense of trust in the system about what needed to be made, and when. As Bestall observes, “It helped expose the difference between what people thought was required, and what actually was required in order to meet our customer service levels.” Another benefit directly resulting from the time savings brought about by Preactor was the ability to remove the need for 1 of the 3 full time planners.
The Search for a Solution: Why Preactor?
A number of factors combined to cause Armstrong to look for a more effective production planning and scheduling system, not least of which was the trend towards increasing product complexity despite a reduction in SKU numbers from 3500 to 2500. For example, whereas there had previously been only one base structure used, there was now a range of 5 possibilities and also a wider range of finishing coatings making it almost impossible for planning to be more than an educated guess. Longhorn had already seen a powerful demonstration of Preactor in action at a regional manufacturing forum in 2004 and was very impressed by its graphical planning board, intuitive use, complete interconnectivity of all processes and the resulting visibility the system offered. As she recalls, “We did look at other competition at the time but nothing offered the combination of benefits that Preactor clearly demonstrated. Specifically, Preactor offered much greater visibility of the impact of one decision on another, which I could already see would be a huge help to us in helping the workforce in the company understand the company as a whole was affected by what they did and did not do.”
Implementation and Go-Live
the board to go ahead with Preactor, Longhorn approached local Preactor reseller RMS after attending a local evaluation session. “It was clear from the outset”, comments Longhorn, “that RMS had a deep understanding of our business, our constraints, and how to make Preactor work best for us to overcome our challenges.” Implementation itself began in August 06 with RMS on site walking through all the business processes with Longhorn. As everything had to be set up from scratch, there was a steep learning curve for all concerned in order to find out not just how Preactor could meet Armstrong’s perceived requirements, but also to identify ways it could help that had not been anticipated. By mid-November Preactor was able to replicate in look and feel the manual schedule familiar to the workforce with the system successfully going live in January 2007.
Silvergate has ambitious plans for the future in the ongoing evolution of its Preactor APS system. These include a move to a total hands off, paper free scheduling process within the company, and extending the use of Preactor APS 400 into like businesses across a number of other sites. Plans are already in process of implementing Preactor in Silvergate’s Italian plant and then running the planning and scheduling of the plant remotely from the UK. As Bestall concludes, “It is our passion and commitment to delivering what our customers want, when they want it, which drives our business. In addition to continually reviewing our internal business processes, Preactor has become an integral part of our ability to achieve our business model.”
Commenting on the early days of using the system Longhorn describes it as a very smooth process. “We had very little need of any help and what was required was only a phone call away with RMS. Any initial resistance to the system was overcome effortlessly within the first few weeks as people could not help but see how Preactor was not only making their jobs better, but also how this was benefiting the company as a whole.” Now Longhorn enters all relevant information directly into Preactor which then generates the schedule for the following week while also allowing much more time to investigate various ‘what if’ scenarios. This allows, as Longhorn says, “a much more inventive approach to scheduling. Whereas before the key objective was simply to come up with a schedule in time to be issued to the plant floor, now we can ensure we are generating the best possible schedule.” Given all the variances that can occur during the manufacturing process, Longhorn manually keeps Preactor updated with any changes to minimise any differences between projected and actual schedule and also to adjust the schedule where required.
The Benefits and Future Plans
The benefits of Preactor were immediately noticed, specifically in terms of delivering the much needed visibility that Armstrong required. With the entire current production plan being visible as well as the ability to stretch the future plan out significantly beyond 1 week, Armstrong could identify areas of waste and processes that could be improved. This has directly had led to a reduction in changeover related WIP from 3 days to under 2, a 33% cost saving. With a move towards enforcing schedule adherence, performance monitoring has dramatically increased which in turn has provided previously unavailable areas of business intelligence about how best to streamline the flow of work through the plant.
These benefits have also enabled Armstrong to begin looking at its planning and scheduling requirements in an increasingly Supply Chain oriented perspective. Specifically Longhorn’s time has been freed up to focus on integrating the company’s internal scheduling with supplier deliveries and also dispatch. Perhaps the greatest endorsement of the success Armstrong has achieved is the decision to investigate how Preactor can help each of the other Armstrong facilities across the globe. Mike Novels, CEO of Preactor concludes, “Armstrong is a perfect example of a global company that has achieved spectacular local results as a result of using Preactor’s production planning and scheduling technology. It also has had the global foresight to see the combined effects of rolling out this technology throughout the group. By doing so, it is following successfully in the footsteps of the likes of Pfizer which also developed its entire supply chain strategy based on Preactor.”
Ardo sees the fruits of visibility by planning with Preactor
Ardo UK Limited, has been a Kent based subsidiary of the Ardo Group since 1994 and supplies high quality frozen fruit and vegetables to over 400 key customers in the Retail and Foodservices sector including Tesco, Sainsbury, Morrisons, Whitbread, Weatherspoons, and Aldi.