The service is based in case study deployed in the London North Western route from London to Carlisle in the UK. The solution utilises edge developments in computational intelligence to analyse railway data for forecasting malfunctions of different components of the railway infrastructure. The solution uses and exploits Big Data to inform decision making in key areas such as cost reduction and efficiency of the rail industry, which affects all stakeholders and rail customers.
For the purpose of the conjoint analysis, the service was decomposed in five attributes:
Table 1. Case study attributes and levels
|Level 1||Level 2||Level 3|
|Value for money||3% increase in revenue||2% increase of revenue||1% increase in revenue|
|Journey reliability||5% less delays in services||3% less delays in services||1% less delay in services|
|Safety||Reduction of accidents with identification of the causes||Reduction of accidents but with no identification of the causes||No overall reduction of accidents|
|Lower cost/price in fee for the infrastructure||Cost reduction by 7%||Cost reduction by 5%||Cost reduction by 2%|
|Reputation||5% increase in travellers/freight||3% increase in travellers/freight||No increase|
- Value for money: Due to better maintenance efficiency, the train operators companies would obtain a better infrastructure for the price they are paying to the infrastructure owner, resulting in an overall increase in revenues in their daily operations.
- 1% increase in revenue.
- 2% increase in revenue.
- 3% increase in revenue.
- Journey reliability: the predictive maintenance system allows the reduction in service delays.
- 1% less delays in services.
- 3% less delays in services.
- 5% less delays in services.
- Safety: the predictive maintenance system results in a reduction of accidents and the identification of the causes.
- No overall reduction of accidents.
- Reduction of accidents but no identification of the causes.
- Reduction of accidents with identification of the causes.
- Lower cost/price in fee for infrastructure: an overall better optimised infrastructure could allow a reduction of the fee payed by the train operators companies to the infrastructure owner.
- Cost reduction by 2%.
- Cost reduction by 5%.
- Cost reduction by 7%
- Reputation: an overall better service results in an increase number of travellers and freight for the train operator companies.
- No increase in travellers/freight.
- 3% increase in travellers/freight.
- 5% increase in travellers/freight.
The results of the conjoint analysis show that the end-users give the most importance to the safety and value for money. They rest of attributes are valued more or less equally with a slight preference for the reputation attribute.
Figure 1. Case study general attributes importance
And after running some market simulations, some of the conclusions reached were:
- A predictive maintenance service counts with great base acceptance given the great benefits it can provide not only in safety but also considering the financial benefits obtained.
- A clear conclusion of the results is that information is one of the strong points of the service, especially security information. Identifying the causes of an accident is valuable information which highly increases the service acceptability.
- Providing reliable estimations about the benefits (given for example the company size or other affecting factors) could be a good marketing strategy.
- In a similar way, it is also worth considering offering the service at a reduced price initially and adjust the price depending on the results obtained during the “trial-period” since the costs derived of deploying the service with reduced functionality are not very high compared with the potential benefits (both for the end-users and the service owner) obtained.
For a more detailed information, see the NEWBITS D3.3 Conjoint analysis on case studies.