19 November 2011
Don't expect many overnight successes when it comes to improving product quality. Annie Gregory discounts magic bullets in favour of hard graft and a holistic mindset
"Defects are not free. Somebody makes them, and gets paid for making them." So said Dr Deming, widely regarded as the father of quality control, nearly 30 years ago. Have we progressed in managing quality since?
Not far enough, according to MAS South East's Nik Brown. "Typically, you'll see quality writ large on mission statements but when you ask questions, you're quickly introduced to their quality manager. It should be up there at the top of the organisation but, more often than not, the person in charge is not even influencing the board."
Quality only hits the top table when there is a real problem and the customer is talking to the MD, or when non-conformance threatens ISO accreditation. Even at plant level, people see quality as owned by quality manager, not themselves. Brown believes they don't see its importance because they don't question its role in the business as a whole: "Is quality taking it forwards in the long term; is it driving value for customers, shareholders and stakeholders? These are the questions that ought to be asked but rarely are."
Brown is another Deming fan: "He knew you have to look at the system as a whole because all parts are interconnected and influence the others." Isolated efforts to improve quality can actually have the opposite effect. "We used to see companies doing two-day kaizen events that would be better called kamikaze kaizen. They typically measured value-added but improvements in one area actually had a negative effect in another. Unless somebody steps back and looks at the whole, the system will fall down."
Brown says he can assess the value of a company's quality management simply by asking how they identify problems' root causes and what mechanisms they have to stop them happening in the first place. "Some still believe they can inspect quality into a business. You can't. It's a contingency, a holding pattern. It doesn't correct the issue. Some are actually proud of themselves for spotting problems in finished goods before they reach the customer. But they don't see the step before – the problem in the process that means it's not capable of delivering what the customer wants."
Nor does he see enough evidence that businesses are trying to align themselves with what their customers really want – one of the first principles of lean. "They align themselves with the textbook factors of quality, cost, delivery, product uniqueness – but not customer choice. What's the complete lifecycle, how does he use it and interface to it? It can be a significant cost to find it out but it means you are planning for success."
Once understood, he believes the priorities volunteer themselves. It isn't always easy to define the critical quality or pin down the voice of the customer. He used to work in the food industry, making Ryvita. He vividly recalls that a burnt taste to some was undercooked to others. But quality processes can still be designed to deliver the unambiguous 'must-haves' like integrity of packaging and total exclusion of foreign bodies.
Knowing customers' real preferences brings a whole new understanding to the value chain. "It involves everyone," explains Brown. "It's about how do we take those orders in the first place, how do we absorb them into the company, which departments do they cross, who delivers the value and how do the customers get that value?" It can be reflected in big issues like completely changing a core manufacturing technique or small ones like removing an unnecessary quality specification. He cites the example of a plastic moulding company with specks of carbon on a part buried deep in the product: "Does it really affect the product or upset the customer? Does it impact product functionality or its appearance? So why is the quality manager spending time getting these parts rejected. There are lots of mini-loops of effort in companies chasing things of no customer value whatsoever." He is equally incredulous that he sees people spending huge amounts of time maintaining processes and policy documents that aren't relevant any more. "Quality managers need the authority to go back to the board and say 'this isn't right – we need to start talking to the customer again, otherwise we are just wasting time within the business'."
When no one understands what the customer really wants, departments work to their own targets, not the process as a whole. "For example, the sales side often over commits because it doesn't understand what the business is capable of doing; it creates systemic failure across the board."
Roland Thompson, previously supply quality assurance manager for Nissan and now with improvement consultancy Sora Group, takes a slightly wider view. He agrees any quality process should reflect business priorities but insists a wide investigation is a prerequisite for narrowing down to essentials. Sora assesses roles and responsibilities, KPIs by company and department, performance against them, overtime and absence levels and a host of other detail with its clients: "We try to find the variables that lead to them missing their quality targets," says Thompson. Their objective is to distil them into a sensible, achievable policy capable of being deployed at every level. (See box p49.)
So many start points, so many issues to address, so many tools available to do it. How does a company with a real, urgent need to change go about it?
When Cy Wilkinson joined Cressall Resistors five years ago as manufacturing director, he knew he wasn't in for an easy ride. The Leicester-based company, which is nearly 100 years old, had just been acquired by Telema and the order book was growing fast. But production was struggling to keep up, quality was slipping and its on-time delivery (OTD) was down to 50%. Today, OTD and right first time (RFT) are in the high 90s even though its sales volume has grown three-fold. What is more, it handles the increased output without any extra floorspace. So what has made such an immense difference? Sadly for the rest of us, there aren't any quick fixes; just a clear understanding of the priorities and a firm strategy for tackling them, using a pick-and-mix approach to virtually every tool in the lean kitbag.
"We looked first at the big issues in the business – the ones that really mattered to the customer. They were obvious – RFT and OTD," recalls Wilkinson. "Our entire operational strategy was based on getting control of our quality and delivery. We studied everything from workforce development to our KPIs, whether we had the right components in the right place at the right time, our factory layouts and equipment, and how it was being maintained." The whole scrutiny was aimed at making sure there were no hidden black holes that quality could slip through.
Cressall makes engineered products to customer-specific requirements. It holds no finished stock. "When all orders are bespoke, most of the room for error occurs at the beginning," explains Wilkinson. "So we put closed loop systems in place to pick them up fast and take immediate preventive action to ensure that when similar projects come up, they won't be repeated." Every instance is logged electronically and can't be closed unless preventive measures are in place and signed off.
Cressall got suppliers on board to manage their own inventory (VMI), introducing kanbans for all fast-moving parts. Major layout changes followed. Optimised flow is obviously difficult when no two projects are the same. Instead it prioritised areas that were critical to turnover and customer service like its Rp coiling section. Output from here goes into 60-70% of finished product yet it occupied less floorspace and had poorer flow than areas contributing 5% of revenue. Investing in a second machine and relaying the area so a single operator could handle both machines opened the bottleneck.
Wilkinson is adamant that quality is an issue for the entire business, not just the shopfloor. So Cressall now has value stream maps (VSM) for the factory as a whole, for each cell and for admin processes as well. It allowed it to cut the 35 steps in its purchase-to-pay process in half and then move it on to an electronic system. It put new, error-preventing workflows into its Efacs ERP system for order entry, approval and issue to production. "On our fastest moving product line, it means we can now produce a bespoke product from standard components by the next day," explains Wilkinson. "That was impossible when everyone was chasing bits of paper around. And VMI and consignment stock mean they always have the components to build it."
VSM also meant that the business, for the first time, had a realistic action plan with clear KPIs – an absolute fundamental in sustaining quality improvements. Wilkinson believes that – amidst all the activity – two techniques made the biggest contribution: 6S (5S plus safety) and visible KPIs. Together, they put organisation, focus, direction – and a fair measure of pride – into the workplace.
Today, there are flat screens in the offices and factory floor which display live output of KPIs from the ERP system. They include orders received as well as RFT and OTD. A daily management board on each section shows its own action plan and KPIs, updated by the team leader. "We spend time in each cell every day to review their plans and where they need help," explains Wilkinson. To make sure the plan for continuous improvement is on target, Cressall has a schedule of regular updates and audits. VSM is revisited for each area every six months, 6S every month, failure mode and effects analysis every quarter.
Today, Cressall's order book is the best it has ever been. Customers come from across the world for a high quality, British product. Of all the tools available for improving quality, what's been most important to Wilkinson? Lean? Technology? "Neither – it's been our people all the way. OK, we focused on lean for the first few years. After that, a big part of my strategy was using IT to build on that. But the real success has come from the people and how well they have worked in teams. Culture and attitude matter most – you have got to have people around you who want to drive the improvements." Even so, he is adamant there is no magic bullet: "We use all the usual tools but quality doesn't come from one thing – it comes when you put them all together."
Keep it simple
Rules of thumb from Sora's Roland Thompson:
- Quality suffers rather than improves when you fire at too many targets. Use your skills and resources to best effect. Focus on what's needed as opposed to what is wanted. Clarify what you are about, pick out some key objectives and develop the strategies to achieve them.
- Keep the policy deployment simple and manageable. The whole purpose is to give focus, making sure everybody pulls in the same direction.
- Understand what's being achieved and why. Manage outstanding issues on a daily basis to make sure people are being transparent. If there's a concern, make sure people aren't frightened to flag it up so you can allocate the right resource and fix it as soon as possible. It's putting round pegs in round holes.
- Communicate. Set a framework where engineering, production and quality can share problems and issues. Make sure all problems are investigated across departments and set up a structure for doing it.
This material is protected by Findlay Media copyright
One-off usage is permitted but bulk copying is not.
For multiple copies contact the