ARMS : سیستم مدیریت منابع خودکار برای PLC مخابرات بریتانیا
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|10471||2006||3 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : European Journal of Operational Research, Volume 171, Issue 3, 16 June 2006, Pages 951–961
Accurate demand forecasting combined with resource planning is critical to a company’s performance and profitability. This paper describes ARMS (automated resource management system), an integrated system developed for the customer service operations of British Telecommunications plc to help with the operational/tactical planning and deployment of the company’s 20,000-strong field engineer workforce. ARMS integrates a forecasting tool with a resource planning tool and a resource balancing tool providing an end-to-end automated resource management solution for the organisation. OR techniques are used throughout the system, including ARIMA for forecasting, constraint satisfaction for problem modelling, heuristic search for problem solving thus demonstrating the value and relevance of OR in solving today’s business problems.
The need for automating resource management, herein referred to as RM, is well recognised and has been the subject of considerable research and development ,  and . The case for automating RM is motivated by the drive to maximise profits, improve quality of service (QoS) and reduce costs. There are three basic tenets of resource management: (i) Forecasting: The ability to forecast the demand of work. (ii) Analysis: The ability to analyse resource and job profiles and identify either over or under resource utilisation. A resource profile refers to a particular collection of location (i.e. area), skill and availability (i.e. time) data. A job profile on the other hand refers to attributes of the job such as priority, start and end dates, job type and so on. Analysis comprises resource planning and scheduling. Resource planning involves profiling where resource profiles are adapted with a view to matching configuration of skills, availability and locations of the resources to the skills, timing and locations of the jobs making up a particular workload. Resource scheduling is concerned with assigning resources to actual jobs and identifying explicit execution times for those jobs. Resource planning is an essential pre-cursor to successful resource scheduling. (iii) Execution: The ability to execute the output of the analysis—in terms of dispatching jobs to resources, making requests for extra resources in case of over resource utilisation. Successful automation of RM requires that the aforementioned tenets be fully automated. Indeed operations research and artificial intelligence methods , ,  and  have been employed to automate some aspects of RM. For example the auto-regressive integrated moving-average i.e. ARIMA model  and  has been successfully employed in forecasting jobs  and . Constraint satisfaction  and  and heuristic search methods  and  have been applied to resource allocation problems. In this paper we describe work we have done in developing an automated resource management system called ARMS, to automate the planning and deployment of field engineers within the customer service division of British Telecommunications plc (BT). BT is the UK’s largest fixed telecommunication services company with over 20,000 field service engineers undertaking repair or provision tasks for the company’s customers on a daily basis and across the UK. BT’s field engineers are allocated jobs via an information system known as Work Manager  and . In order for BT to best serve its customers, resource managers within the organisation have to ascertain how best to plan and deploy the company’s field engineering workforce on an operational/tactical basis (i.e. for the following day and up to 14 days ahead). Generating the plans involves: • Forecasting demand for several activity types (e.g. provision, repair, maintenance work). • Planning the volumes, skills and geographical locations of engineers required in order to service that demand in the next 7 up to 14 days. • Deploying field engineers for tomorrow in the best possible manner so that incoming and existing work can be optimally scheduled by the company’s automated workforce scheduling system  and . Each resource manager is typically in charge of around 150 engineers and he/she has certain degree of control in deciding the planning and deployment of the workforce. For example, the following possibilities can be considered and decided upon: • Engineers are multi-skilled and they can perform several tasks requiring different competencies and capabilities. There are currently nine different “skills” defined for the purposes of RM with each engineer having one or more of these skills depending on past training and/or experience. The resource manager can focus specific engineers to work on activities of a particular skill or type. • Engineers have the flexibility to move around between relatively small geographical areas called “patches”. A customer service team (CST) usually is responsible for several patches and the resource manager has the responsibility to resource each patch adequately by moving engineers from neighbouring patches or by altering their working pattern to be explained next. • Engineers have different working patterns. They can work full day or half a day. They may also be available to come and work on overtime on voluntary basis. Furthermore, they have several business related absences or meetings scheduled which, if necessary, could be cancelled to increase availability in certain geographical areas or on specific dates. Considering all these factors, generating a resource plan means forecasting demand in terms of time, skill and location in addition to finding an optimal configuration for all the resources and that over several days. Sub-optimal solutions are bound to be generated setting aside the substantial effort required in producing them manually. This is to be expected since, even if we set aside the forecasting requirements, the core problem is an NP-Hard optimisation problem and can be considered as a variant of the integrated location-routing problem , which is known for its complexity. The breakdown of the remainder of this paper is as follows. In Section 2, we present the forecasting tool of the ARMS system. In Section 3 we describe the resource planning tool also called Dynamic Planner. In Section 4, we provide an overview of the resource balancing tool called Collaborator which sits on top of Dynamic Planner and it is used when managing resources over large geographical areas. Section 5 provides a summary of benefits and concludes the paper.
نتیجه گیری انگلیسی
We have presented ARMS, a system for automating BT’s resource management. The three market drivers—globalisation, deregulation and technological innovation—have highlighted the need to move from reactive to proactive resource management. Proactive resource management calls for a positioning of one’s resources so as to service jobs in an optimal manner. Parts of the ARMS system are fully operational and deployed while others are currently under pilot in different parts of UK. When the system is fully deployed, it will be interconnecting and cross-optimising resource management activities throughout the company. This will make the enterprise more responsive to market demand, able to streamline its operations and offer a seamless end-to-end customer service process. Some of the quantifiable benefits from ARMS can be separated in those resulting from the Forecasting tool, those resulting from Dynamic Planner and also overall benefits. In the case of Forecasting, the benefits can be summarised as follows: • The tool automates the workstack forecasting process which was previously manually performed using spreadsheets. • Standardises forecasting practices across the organisation allowing an auditable process to be setup which can be effectively monitored and improved if required. • Improves forecast accuracy over the prior locally utilised methods. Our test data has shown that the MSE (mean square error) could be improved by adopting ARIMA up to 64% for the repair workstream and 14% for the provision workstream. In the case of Dynamic Planner, the benefits can be summarised as follows: • Dynamic Planner automates the resource planning process within the company. The projected cost savings only from process automation are estimated to be up to £1 million. • It standardises resource planning practices across the organisation with similar benefits to Forecasting on that respect. • It optimises deployment of resources by increasing workforce utilisation and mobility. This results in improving our quality of service while reducing operational costs. • It substantially reduces overtime costs which currently are estimated to be up to £18 million/annum. BT aims to reduce substantially its operational costs of employing a workforce of 20,000 engineers from using ARMS. A study is planned to fully assess this impact but operational savings of between 1% and 5% per annum could be expected which, given the current overall costs involved, could translate to savings between £10–50 million/annum for the organisation. Substantial customer satisfaction benefits are also expected because of the better appointing of work and resource availability where it is required by our customers. Overall, ARMS has excellent prospects and there is currently interest to introduce the system to other parts of the company such as our Wholesale division.