Embed Size (px)
Citation preview
Service Systems Engineering: Lessons Learned from Case Studies
Stephen J Ashlin, QinetiQ,
Iain Cardow,
Rolls-Royce plc, [email protected]
Alan Crawford,
Babcock International Group, [email protected]
John K Davies, University of Leeds,
Andrew Farncombe, [email protected]
Edwin Swindenbank, Harmonic,
Copyright © 2016 by S J Ashlin, I Cardow, A L Crawford, J K Davies, A Farncombe, E Swindenbank. . Published and used by
INCOSE with permission.
Abstract. This paper examines and compares two Case Studies of the development and operation of
Services:
• The introduction of Electric Vehicles through rental companies in a holiday area of Japan
• The introduction of the Transport for London Cycle Rental Scheme.
Both these services appear novel but in different ways. The operation of car rental schemes is well
understood; the main differences here appears to be the technology used. The TfL cycle scheme is
based on established technology; its novelty is in its operation and its user base of London
commuters. The case studies have been examined in order to identify what was good and bad
about the services, to compare with and validate the findings of earlier work and explore if other
areas need to be considered. One service can be seen as a failure, whereas the other has been a
success.
This work reveals similarities and differences which may be apparent to the reader in a general way
but can be clearly shown through the framework provided by our approach.
The major findings of the work are that the structure of Service Definitions is useful in understanding
the service, however in one case it was not clear who the Service Customer was for the overall
service and this led to major issues. The consideration of the Characteristics of the Case Studies
identified critical characteristics that had not been considered within the Case Study, and two of
which had not been identified through our earlier work. A more detailed use of characterisation is
important in identifying key areas. However, in addition by comparing the two studies the
importance of the interaction of stakeholders with the service, in particular the Users of the service,
is key to success or failure. These aspects have not had sufficient attention in our previous work and
need to be included in future recommendations.
Introduction The INCOSE UK Service Systems Engineering Working Group is working to identify areas of the
engineering lifecycle, methods, techniques that need to be different for delivering services as
opposed to ‘traditional systems engineering’ for delivering products. The overall aim is to provide
guidance for the development and operation of services by our members whether acting as Service
Providers or Service Customers.
Out previous work in this area (Ashlin et al) considered ways of identifying what makes particular
services difficult to develop based on a range of categories covering Stakeholder, Management and
Technical aspects. Existing definitions of Service were analysed and definitions of entities related to
Services used to provide a means of understanding different definitions and of scoping the concepts
for services. Architectures for Services were considered and the use of Service Oriented
Architecture, SOA, seen as a way forward.
Reaction and comments on this previous work has led to further investigation of the wider field of
Characterisation of services and Architecture Frameworks through the use of Case Studies.
This paper examines and compares the results of two Case Studies:
• The introduction of electric vehicles through rental companies in a holiday area of Japan
(Weiller and Neely)
• The introduction of the Transport for London Cycle Rental Scheme. (TfL Cycle Scheme)
The Case Studies use information available in the public domain but here are considered in terms of
the services they provide and the approaches covered in our earlier paper.
Approach
Overview
The two Case Studies are described using text from their respective references. The major issues are
identified and discussed and areas of success or failure are identified. The Case studies are
compared with the Service Definitions, the Service Characteristics and the Service Oriented
Architecture presented in our previous work.
OKINAWA Electric Car Rental Scheme
Background
The search for alternative fuels to reduce car CO2 emissions is an important part of the climate change challenge. Electric vehicles (EVs) are one potential solution. Yet the EV market has been slow to develop. The global stock of electric vehicles was some 180,000 vehicles at the end of 2012, just 0.02% of total passenger cars.
In the early-stage EV market, electric car rental services allow drivers to trial an EV for a short time. This experience with EVs increases awareness and facilitates diffusion of the technology. That is why projects like the Okinawa Electric Vehicle (EV) Rental Service are so significant. One of the first EV projects of its kind in the world, the service is dedicated to supporting the use of electric cars and improving the environmental sustainability of tourism on the island of Okinawa, Japan.
Unfortunately, the results of the Okinawa EV Rental Service at the end of its first three-year operational phase in 2013 have failed to meet initial expectations. Low usage rates mean rental companies are making a loss. Customers worry about insufficient recharging infrastructure. Sales of used rental cars are low.
There are potential solutions to these challenges, however. New research offers fresh insights into the scheme. By adopting changes to the strategic approach – in particular, harnessing the resources in the business ecosystem and sharing the data – it may be possible to spark new life into the venture.
The Service
The introduction of electric vehicles to Okinawa through the holiday rental service was the first phase of a three-part “Green New Deal” aimed at developing a smarter energy system on the island. A combination of factors makes Okinawa a suitable location: high rates of car ownership, lack of public transportation, and its popularity as a tourist destination – welcoming 5.5 million visitors per year, of which half rent a car.
Originally proposed by professors Hideaki Miyata and Kenji Tanaka from the Department of Systems Innovation at the University of Tokyo, the EV rental service complements the aim of local organisations to improve the environmental sustainability of tourism on the island.
A broad network of organisations participates in the service. Researchers at the University of Tokyo designed the business model and conducted the technical and market research. Travel agencies across Japan are the service’s main distribution channel, as they offer package holidays that include electric vehicle rentals.
Three rental service companies provide 200 EVs between them: Nippon Rent-a-car Okinawa, Nissan Rent-a-Car Okinawa, and ORIX Rent-a-car Okinawa. The cars are Nissan Leafs with a range of 160 km. Car hirers either pay to use the 27 fast-charging stations on the island or recharge for free at a slow-charging station, which takes about eight hours for a complete charge.
The fast-charging infrastructure for the EVs is provided by Advanced Energy Company (AEC), a joint venture of 26 companies. Other members of the EV Rental Service ecosystem include the Shiraishi Group, a major financial holdings company, local and federal public authorities, and local companies and tourist site operators that offer slow-charging stations.
Business Case
For car rental companies, acquiring a new Nissan Leaf costs about 3M yen. The rental company expected a resale price of 2M yen ($20,000) after three years of rental. To break even the rental firm must generate at least 1M yen over a three-year rental period. At typical hire prices of 6,300 yen/day, that equates to approximately 18 rentals per car per year, assuming an average rental lasts three days. The target was 100 rentals per car per year.
AEC’s business model is based on providing access to the network of fast-charging stations in Okinawa. Regulations governing electricity sales in Japan forced AEC to position itself as an infrastructure provider charging a flat fee, rather than an electricity provider selling kWh. AEC charges a service fee of 2,000 yen ($20) for unlimited recharging during the EV rental. After a week the service is charged according to the number of times stations are used. The business model was based on an average distance driven per rental of 250 km and the assumption that a fully charged Nissan Leaf would only need one full recharge per average three-day rental.
Challenges
Estimates of usage rates in the original business plans proved to be optimistic. The EV usage rate reached 10.6% in 2012 and 20% overall since the start of the project – corresponding to 60 days usage/car/year.
The car rental companies met with a lack of demand for used Nissan Leaf rental cars in the local market. The average income on the island is too low to create an EV market, even for used cars. Consumers are only willing to pay around 1.5 M yen ($15,000) per car rather than the 2M yen anticipated by the rental companies.
Customer feedback suggests “range anxiety” is a major factor in the low adoption rates. Sixty per cent of customers were anxious about the number of charging stations and running out of fuel. Fast-charging station coverage was seen as too sparse when compared with gas stations. Consequently, tourists tended to recharge their hire car’s charge two or three times a day, rather than once as calculated in AEC’s original business plan. Customers also said that the service was not sufficiently price-competitive with conventionally fuelled car rentals to justify the risk.
Another significant challenge is the lack of incentives for travel agents to promote the use of the electric vehicle option over conventional vehicles. Because EV technology is less well-known, travel agents cannot guarantee the performance of the vehicle beyond existing official statistics about range, safety, and ease of charging. Also, they are unable to respond to concerns about the availability of charging on Okinawa.
Finally, the Fukushima nuclear reactor crisis, triggered by the March 2011 earthquake and tsunami, negatively affected consumer sentiment about the electricity industry and related innovations such as electric vehicles. Since 2011 energy infrastructure and supply companies have reduced focus on electric mobility in Japan.
Results
Why was the Okinawa Electric Vehicle Scheme perceived as a failure?
Service composition: The Service Users are the holidaymakers who take out standard contracts with
the Car Rental companies. The rental companies buy the cars in from Nissan and plan to sell them
on to the local population after three years use. The main difference for these cars was the
provision of the recharging stations at the convenience stores. The Service Architecture and
Components are all well-known and relatively simple.
Study: No detailed study of Service User’s needs and concerns. Assumption that ‘going green’ was
sufficient to create the need.
Information: Lack of information about Range and Performance for potential Service Users. Leading
to reduced rentals and queues at Recharging Points. This led to more rechargers being needed, but
the funding not being there to provide them. Lack of trade to the Convenience stores as customers
avoided those with queues in the car-parks.
Co-operation: There was lack of co-operation between the various stakeholders and lack of
awareness of the Service Provider (the Car Rental Companies) about what was actually happening.
Lack of discount from the Manufacturer to the Car Rental companies prevented them from reducing
charges for EV rentals. Lack of incentives from the Car Rental Companies to the Travel Agents so
they would promote the electric cars over petrol cars.
Environment: There was a lack of awareness of the environment into which the Service was
deployed. Hotel staff are paid below average wages, so the price in the sell on market was less than
planned. The Fukushima nuclear reactor crisis negatively affected consumer sentiment about the
electricity industry.
Feed-back: The feedback from initial users to further users was bad, leading to 30% of the electric
cars being rented out at any one time, compared with 90% planned.
Comparison with Service Systems Engineering Approaches
Figure 1 Service Definition Map
In terms of the Service Definition Map, the Contract covers the Rental Agreement between the
holiday makers and the car rental companies. The holiday makers are the Service Customers, the
Rental Companies are the Service Providers and the Travel Agents the Service Brokers. The Rental
agreement forms the Service Agreement, however, car rental agreements are more about putting
conditions onto the Service Customer than the Quality of Service provided by the rental companies.
Other stakeholders provided further parts of the overall service. However, it is not clear if there
were clear and documented agreements between them. The Car Rental companies were the
providers of the overall service, they were the main ones who lost money on the exercise, but they
do not appear to have taken the overall lead on the programme. The Business Case was fairly
standard for a Car Rental company, but produced by Tokyo University. When things started to go
wrong, the Car Rental companies did not go back to the business case to consider what needed to
change in order to improve things.
The Service Specification does not appear to have been considered at all. The change to electric cars
has been regarded as a small addition to normal business with regard to the Charging Stations as
part of the provision and operation. The high level aspects of Specification, Provision and Operation
have not been considered.
In terms of Service Characteristics, when the wider aspects of the service are considered, (rather
than just those concerned with renting the vehicle), many of the standard characteristics apply to
this service.
Figure 2 Characteristics of Services
However, there are some Characteristics that emerge from this Case Study that have not been
captured through previous work. Within the Management area, the Risk of lack of take-up of the
Service and the overall commercial risk (who loses money when it fails) had not been identified. In
the Stakeholders area, there was a general lack of ownership of the service and making sure it would
be a success. The lack of communication between stakeholders was a major issue and contributed
to the failure. In the Technical areas, the Service Requirements, Design Transition and Operation are
missing, and in a general sense the lack of Information about the system needed by other
stakeholders (such as the vehicle range and performance) contributed to the failure.
Figure 3 SOA Model: Electric Car Rental
With regard to the SOA model, it is relatively easy to construct a business process model and relate
the services to the activities and how they are provided. However, what this does not address is the
dynamic operation of the service, both in terms of event driven aspects ‘what happens when?’ and
performance aspects such as provided system dynamics models. Such models are needed at many
levels to study the operation of the service at the business level, the operational level and the
service level. In addition, there are a whole range of non-technical issues that are not covered by
the SOA model but should be taken into account.
Transport for London Cycle Rental Scheme
Background
In 2007, the city of Montreal published the "Reinvent Montreal" transportation plan, which featured
a bike sharing system as a method of reducing automobile dependence. To design and operate the
new system, the city created the Public Bike System Company (PBSC), a private non-profit company
overseen by Stationnement de Montréal, known as the Public Bike System Company, serving as the
Montreal Parking authority. (BIXI)
Similar services have now been provided in further cities in the US, Canada, UK and Australia. A
complete station is made up of a pay station, bikes, and bike docks (where the bikes are housed),
which are fitted into modular technical platforms (racks) that are powered by solar panels. These
technical platforms are the base and electronic ports for pay stations and bike docks. Bike stations
can be created, expanded, configured and removed in about half an hour, monitored by a real-time
management system. Excavation or preparatory work is not required, enabling the installation of a
bike station as an adjunct to on-street parking.
The Service
In 2009 Serco were awarded a new contract by Transport for London (TfL) to design build and operate the London Cycle Hire Scheme (LCHS). which went live in 2010. (London Cycle Hire Scheme Agreement) The scheme provided over 6,000 bicycles for use in central London supporting over 50,000 journeys a day. Hiring charges are £2-00 per day, which enables the cyclist to travel any number of journeys of 30 minutes or less with an additional £2-00 fee for each further period of up to 30 minutes. (Yes you can travel all day for £2-00 as long as each segment is less than 30 minutes). There are over 400 Racks with over 10,000 docking stations where bicycles are stored, hired and returned. Payment is automated through use of credit cards. The scheme is part of the TfL transport initiative and links into the TfL travel website which provides information on routes from the user’s current position to a specified location by the various modes of transport, cycle and rack availability, route options, and real-time route guidance to mobile phones using the phones SatNav. The scheme was specified by Transport Trading Limited, TTL, a subsidiary of TfL who carried out
extensive studies into user population and expected use. (Cycle Hire Communications Development
2010) The Service Provider is SERCO the leading service company in the UK . (Serco)
The scheme has been a major success and has been extended to provide 9200 bicycles and 687 racks.
Results
Why was the TfL Cycle Scheme a success?
Service composition: The Service was (and is) provided and operated by Serco using components of
the Service utilising the bikes and racks provided by BIXI in Montreal and their suppliers. The Service
Architecture and Components are relatively simple.
Study: Extensive study of potential use and users, commuters and locals. Identification of User
population, concerns and benefits, leading to the number and locations of racks and bikes and the
payment and pricing policy.
Information: Real-time provision of on-line information to the Service Users for planning, routes,
available bikes and racks, integration of information with Rail, Tube and Bus Services, and SatNav
route guidance on mobile phones.(TfL Route Planner)
Co-operation: The Service Provider and Service Customer worked together to determine the level of
specification of the Service and extensive specification of Service User related aspects. Quality of
service for maintenance, repair and replacement of Bikes and Racks, based on Service User reporting
when returning bikes. Monitoring of bike/rack usage and rapid adjustment of bike location when
needed.
Environment: The environment into which the Service was deployed. Congestion Charges leading to
reduced motor traffic in central London. Increasing use and congestion of rail, tube and bus
services. Climate allowing year-round usage of the system.
Feed-back: The feedback from initial users to encourage further users. Realisation that bicycle theft
of Users bikes was no longer an issue.
Comparison with Service Systems Engineering Approaches
In terms of the Service Definition Map, there are two levels of Contract. The cyclists are the Service
Customer and they in principle have a Service Agreement with the Service Provider, who can be
thought of a Serco. However, in commercial terms, TfL collect all the money from the cyclists and
have a Service Contract with Serco to provide the Service Operation, so they act as Service Provider
to the Cyclist and Service Customer to the Service Supplier. There is no identified Service Broker.
There is a Service Agreement and Service Level Agreement between TfL and Serco for the provision
of the Service. The Service Specification and Service Interface exist but the concept of Service
Catalogue and Service Line are not useful in this Case Study.
As Service Provider, Serco is responsible for the Service Provision and Service Operation, providing
the activities within that group.
In terms of Service Characteristics (Figure 2), in this Case Study the wider aspects of the service have
been considered during the Service Definition Phase. Of particular note are the Reliability and
Availability of the Service, which includes how many cycles are currently available at each rack.
Flexibility of Scale the service capable of future expansion in all its dimensions, such as numbers of
bicycles, racks, and locations. The Management and Risk areas are relatively well covered, but the
Case Study. However, there are some Characteristics that emerge from this Case Study that have
not been captured through previous work. Within the Management area, the Risk identifies the
issue of lack of take-up of the Service and the overall commercial risk. In the Stakeholders area, the
ownership of the service and making sure it would be a success was clearly seen. The
communication between stakeholders was good and the provision of technical information about
the current use of cycles and racks being available to the Customers via the journey planning
applications was a major factor in the success of the system.
Figure 4 SOA Model: TfL Cycle Scheme
Again with the SOA model, the business process model can be constructed and show the services
related to the activities and how they are provided. However, there is again a large area of the
modelling that this does not cover that is needed for the successful specification design and
operation of Service.
Conclusions Comparison of the two Services has been made using the success and failure factors identified
above. In terms of complexity, both services are relatively simple in terms of components. For the
Electric Vehicle scheme, similar rental schemes exist across the world. The difference here was the
mode of power and the necessary infrastructure and chargers.
A major difference between the two schemes is in the amount of work done in the Service
Specification phase. Very little work was done for Okinawa. It appears to have been largely
assumed that as the service provided by electric cars was the same as that provided by petrol-driven
cars, there would be little work to be done changing from one to the other. On the other hand,
extensive work was done for the Cycle Scheme in terms of identifying Stakeholder Needs and
developing models to understand usage and operation.
Table 1: Summary of the Comparison of Findings
Electric Vehicle Scheme Cycle Scheme
Service Composition Simple – mostly existing Simple – existed
User Study None – believed it would
‘sell itself’
Extensive study to identify
needs
Information No information on Range
and Performance
On-line information on Routes
and availability
Co-operation Lack of co-operation
between stakeholders
Good co-operation between
stakeholders
Environment Holiday staff poorly paid Need for more cycle journeys
Feedback Bad – reduced uptake Good – led to expansion
Provision of Operational Information to the right people at the right time has proved crucial in both
these Case Studies. The lack of information on range and performance severely limited the use of
electric cars, whereas the provision of current status information to cyclists greatly increased the use
of cycles and the efficiency of the service.
The recognition and co-operation between stakeholders can be seen as a major factor in success one
service, and its absence can be seen as a cause of failure in the other. In particular, the transition to
service needs to be well designed and managed.
Making assumptions about the environment are dangerous, running car hire in Okinawa is different
to that averaged across the country; what works in London may not work elsewhere.
In terms of overall process, the differences between the two Case Studies can be understood in
terms of the application of good Systems Engineering to the specification, functional design,
implementation and operation of the service.
Usefulness of the Service Systems Engineering Approaches
The Service Definition Map has been useful in identifying what needs to be considered in delivering a
service and where the lack of definition causes issues and failures. This map should be formalised to
produce guidance on breakdown of service definitions.
The use of the Characterisation of Services identifies which characteristics are important for a
particular service. Some previously identified characteristics have been seen as important and
further characteristics have been identified. These mostly relate to the Service Specification phase
of the lifecycle. With regard to Risk, the risk of the service users not wanting to use the system and
the overall usage being low was not considered in the Okinawa Study, but was clearly identified and
addressed in the TfL Cycle Scheme. This aspect had not been included from our previous work.
In the Stakeholder area, the exchange of information and co-operation between stakeholders was of
major importance in Okinawa, but did not happen. In TfL it was seen as crucial.
Information was key to success or failure in these use cases. It forms a major part of the TfL Cycle
Scheme, but was lacking in Okinawa and cause a lot of the issues.
Overall the list of Characteristics was important in studying Services but different characteristics will
be important in i providing different services
The Service Oriented Architecture is important in visualising the overall Business Process, the
services needed to provide that service and how they are provided. However, it is a static,
qualitative model and does not provide any assistance in evaluating the performance of the service.
Further techniques such as Influence Diagrams, Systems Dynamics (System Dynamics Society) and
simulation tools (Liu et al) are needed to evaluated the performance of the service. These include
early modelling to ensure the service will meet the Service Level Agreement (Liu et al) as well as any
service requirements.
On the basis of the work presented in this paper, it appears that Systems Engineering, (ISO 15288),
(INCOSE Systems Engineering Handbook), applied properly, is needed for the successful
specification, design, delivery and operation of services.
Further Work
The use of these Case Studies has driven forward the work on Service Systems Engineering and
brought it closer in line with the use of Systems Engineering lifecycles and techniques. The findings
will allow development of guidance for the engineering of services to be made available to INCOSE
members.
The two Case Studies were partly chosen as they had similarities but different outcomes from which
lessons could be learned. However, because they have similarities, (personal transport, green
agenda) a wider range of further Case Studies that appear to present different business domains,
scales, complexity and problems should be studied. The use of two case studies in a similar domain
area has been useful – as they have similarities and striking differences. Currently available Case
Studies are mostly from those in the public domain. To provide guidance that is more specifically
directed to the work of INCOSE members it would be good to have more Case Studies directly from
their own businesses.
References
Ashlin S J. Understanding Services: Understanding Stakeholders. Paper presented at the 2016 INCOSE International Symposium
BIXI Société de Vélo en Libre-Service. https://en.wikipedia.org/wiki/Bixi_(company) Accessed 7th June 2016.
Cycle Hire Communications Development 2010. http://content.tfl.gov.uk/cycle-hire-communications-development-presentation.pdf Accessed 7th June 2016.
London Cycle Hire Scheme Agreement https://cclondon.tfl.gov.uk/corporate/publications-and-reports/cycle-hire-contracts Accessed 7th June 2016.
Liu L; Russell, D.; Webster, D.; Luo Z; Venters, C.; Xu J; Davies, J.K., Delivering Sustainable Capability
on Evolutionary Service-oriented Architecture. Ob-ject/Component/Service-Oriented Real-Time Distributed Computing, 2009. ISORC '09. IEEE International Symposium, pp.12-19, 17-20 March 2009
Serco. https://www.serco.com/media/pressreleases/londoncyclehire-843Accessed 7th June 2016
TfL Route Planner Web Site/App https://tfl.gov.uk/plan-a-journey/ Accessed 7th June 2016.
TfL Cycle Scheme https://tfl.gov.uk/modes/cycling/santander-cycles Accessed 7th June 2016.
Weiller C. and Neely A. Cambridge Services Alliance Electric Vehicle Rental Services: Project in Okinawa, Japan. http://cambridgeservicealliance.eng.cam.ac.uk/news/Okinawa%20Report Accessed 7th June 2016
System Dynamics Society. Introduction to System Dynamics http://www.systemdynamics.org/what-is-s/ Accessed 7th June 2016.
ISO/IEC/IEEE 15288:2015. Systems and software engineering -- System life cycle processes
INCOSE Systems Engineering Handbook v4.
