London Underground

LU challenged SISCOG to develop a tool that needed to be fast and flexible at scheduling and capable of optimising the work of the Customer Service (station staff), just as it had done before for the drivers/train operators (LU has been planning and optimising the duties of around 3500 drivers with the CREWS-based system TSS since 2008).

But LU needed to be sure that their needs and goals would be perfectly met. So, a proof-of-concept (PoC) was developed upfront and convinced decision-makers that this was the way to go to achieve an optimised feasible work plan.

The goals laid out for the around 5000 station staff scheduling system were to:

• originate an optimised scheduling solution where staff preferences were maximised
• produce a valid and efficient set of duties and rosters
• present a faster and more flexible planning process
• provide autonomy regarding the definition, management and testing of labour rules
• support as many groups of station staff as possible, with guaranteed scalability in number and complexity
• provide a less error-prone and swifter User Experience
• generate outputs to SAP systems
• generate key reports

SISCOG answered LU´s challenge with the development of a system, based on the Duty Scheduler and Roster Scheduler modules of its CREWS product, and aligned with the 2008 mature brother TSS - Train Staff Scheduling system for drivers.

The project approach was based on an “add value along the way” philosophy and dealt with as a partnership where there were not two typical client-supplier teams, but just one team. Developed through an agile methodology, this allowed for a constant alignment between SISCOG and LU, with 2-week interactions and releases every-other interaction, fostering the possibility for early use of the system by the end-users as a source of right-path confirmation and motivation.

On top of all the proven capabilities of the CREWS product, the system incorporates a new feature – the Rules Editor. This editor injected the necessary flexibility that LU needed to rethink the current Customer Service organisational structure. And be capable to continue to do so at any time, by the possibility to configure the rules that will let the system know how to validate and optimise duties and rosters. SISCOG's optimisers use this information to guide the optimisation process, which really gives the end-users all kinds of possibilities. With the activation of the Rules Editor, LU specific rules and staff preferences are all expressed in a Rule Book, as is the optimisation’s engine cost rationale, leveraging the user experience to a whole new level.

Detailed training sessions and a final workshop enabled the main users to get the most out of all the system’s functionality and train others.

The system is being used and under maintenance since 2022 enabling LU to rethink Customer Service, while catering for station staff preferences in an optimised manner.

• Important savings in the new work schedules
• Maximised station staff preferences
• Decrease from 35 to 15 hours to plan one large complex station staff group
• Autonomy to redefine the organisational structure of the Customer Service
• Automatic validation by the system regarding the feasibility of duties and rosters
• Management of business knowledge in the system
• Better scheduling decisions due to the capability to build “what if” scenarios