Banking crisis

Tube Lines recently embarked on a project to accurately measure earth structure movements on cuttings and embankments along the Jubilee, Northern and Piccadilly lines. Eva Holmes and Nick Whelan explain how Lean Six Sigma techniques helped to identify savings and create a new benchmark for the transport and engineering industry.

Tube Lines has the long term responsibility for the maintenance, renewal and replacement of all assets on three of London Undergrounds (LU) busiest lines, including 255 trains, 100 stations, 320km of track and tunnels and 227 escalators. We are also responsible for the 80km stretch of trackside cuttings and embankments.

Within the Tube Public Private Partnership (PPP) contract we have clear condition benchmarks to meet for cuttings and embankments which require us to assess each structure to establish their condition and then carry out remediation work where the condition of a structure is found to have deteriorated.

By mid 2010, which marks the end of our first review period when economy and efficiency are judged by an independent arbiter and the needs of London are reconsidered, we are required to tackle sites with a ranking of E and D and, in some cases, sites with a C ranking.

Since we started back in 2003 we have carried out assessments on 2,000 sites of which 15 per cent have shown a decline in condition rating and about 41 sites have been re-mediated.

However, a review of our assessment process, which started back in march 2008, highlighted problems with the techniques used to measure the slopes. In short, the measuring techniques using a simple tape measure and compass were not entirely accurate and gave us readings that were only 25 per cent reliable. The result? We were carrying out unnecessary remedial work on slopes and the cost of this poor quality (COPQ) was over 1.5 million.

Our contract with London Underground (LU) requires us to carry out improvement works in an economic and efficient manner. This was clearly neither economic nor efficient. So we set about challenging this accepted industry practice to find a more effective way of carrying out our improvement programme. To that end, we turned to Lean Six Sigma.

Six Sigma is a business management strategy originally developed by Motorola in the mid-1980s as a statistically-based method for reducing variation in electronics manufacture. It has since evolved to encompass many different elements of business improvement, including sustainable improvements to business processes. It uses a set of quality management methods, including statistical methods, and creates a special group of people within the organisation who are experts in these methods. Such experts are known as Black, Green or Yellow Belts, depending on their level of expertise.

Today we have nine trained Six Sigma Black Belts whose role it is to review the accuracy and quality of work being undertaken by the business with a view to improving business output. We use it as a process improvement tool which is reaping tangible benefits, including reduced project cycle times, lower project costs, improved quality and better customer satisfaction levels. In 2008 our Six Sigma programme generated 36 million benefits, of which 12m were hard savings.

Rooting out the problem Given that remedial work was being undertaken on slopes that didnt require it, our objective was to reduce the cost of maintenance per metre of earth structure by just over 1.5 million. We aimed to achieve this by October 2008 when the next annual round of inspections was due.

Using various Six Sigma techniques, including process mapping, data analysis and failure mode effects analysis, we began by reviewing historical data on 36 slopes, examining the length, angle and height of each slope over a 13 year period. What we found was that, over the years, the slopes appeared to have varied in length, height and angle. For example, on data collected from one slope the angle had changed by seven degrees, the height by seven metres and the length by 14 metres. We would normally expect an earth structure to move less than 30mm in any one year, except if there was an issue with stability. To find extensive movements in 36 slopes like this over a short timeframe didnt ring true.

On analysis we found that one of the problems in data collection was the reliance on track marker plates. Track marker plates are positioned along a slope at 100 metre intervals to mark the location of the inspection point. However, either during routine maintenance or through general wear and tear these plates are removed and replaced, and often put back in a slightly different location. The impact of this is that when the slope is subsequently measured it is done so from a different starting point, thereby providing data variations and inaccuracies.

Furthermore, and according to LU standards, we can measure the slope five metres either side of the axis. This gives us a ten metre range from which to start the inspection giving a greater potential for error and variations of measurement. For example, if a track plate is missing and our pacing is 10 metres out then there is a strong possibility that we are measuring 20 metres from the actual intended point of inspection. This will clearly impact the angle, height and length measurements of a given slope and this could be mistaken for slope movement where there is none.

The use of basic measuring tools also contributed to the inaccurate measurement of slopes. The industry norm is to measure slopes using a tape measure and a compass with an inclinometer. A tape measure can easily get caught up in surrounding vegetation which could affect the measurement. The inclinometer is about 80mm in length which is clearly insufficient when measuring the angle of a slope around 20 metre in length.

The tape measure and inclinometer combined was only 16 per cent accurate for measuring height and angle, and only 25 per cent capable of measuring length when compared to the new solution.

(Above: Award winners - Improvements in measuring techniques were recognised by the British Quality Foundation Lean Six Sigma Academy, which named Tube Lines winner of its Project Excellence 2008 award.)

Finding the solution

We replaced this outdated measuring technique with a global positioning system (GPS) which uses Russian and American satellite systems and provides measurements within 30mm of accuracy. It also provides the height, length and angle of the slope without the need for additional calculation and can measure multiple slopes at any one time. The data is stored on the device and immediately transferred to our internal systems for interpretation and storage.

The benefits of this new measuring technique are obvious; the time to accurately measure a number of slopes is significantly reduced and in most cases halved, and the time to transfer the data from the field to computer for analysis is also reduced. With data automatically transferred to internal systems the risk of losing data is also removed.

To test how more effective this GPS tool is we undertook a sample of measurements from a slope using the traditional measuring technique and compared the data collected from the GPS equipment. Over four readings, the old technique produced variations of 2,110mm compared to just 120mm with the new and improved system. With the new system, variations in measurements have been reduced by 95 per cent, giving our engineers the confidence that they are working from accurate data. If the data shows any variation in measurements then they can confidently judge that further inspection and potential remediation is required.

Bottom line, award winning results

The implementation of this new method has led to a potential project saving of 2.7 million, far beyond the 1.5m target we set ourselves.

In October 2008 we re-started our inspection programme using the GPS equipment. Our asset engineers are confident that we will now be able to target our remediation of earth structures where it is actually needed, thereby reducing the cost of maintenance per metre of slope. This in turn means we will reach our contractual benchmarks in a more efficient and economic manner.

• Eva Holmes is the Six Sigma deployment manager in Tube Lines, managing an extensive Six Sigma programme delivering 50+ process improvement projects annually. Nick Whelan is involved with the maintenance of the Jubilee, Northern and Piccadilly lines. He became a Six Sigma Black Belt in early 2008.