Spatially Enabled Solutions: A Modern Approach to an Old Problem

Spatially Enabled Solutions: A Modern Approach to an Old Problem

Frederick R. Pinkerton
F. R. Pinkerton & Company
275 13th Street, N.E., Suite 402
Atlanta, Georgia 30309-3698
Email : frp@frpinkerton.com

Abstrct

As communities have grown and changed over time, traffic patterns and geographic concentrations of utility fieldwork have also changed. These changes typically create inefficiencies in otherwise well-planned field "networks" of service center facilities, crews, vehicles, and assigned territories (service delivery networks). An outdated or inappropriate service delivery network (SDN) configuration causes unnecessary field-force windshield time, which can significantly affect productivity, response time, and overall cost. Too many service centers or inappropriate staffing levels cause unnecessary facility and labor costs. Too few service centers or inefficient territory assignments increase windshield time and travel costs, and decrease responsiveness and field-force efficiency. Updating the SDN configuration to address current and future work demands has long been a challenge that previously was difficult, if not impossible, to effectively approach. Fortunately, improved customer, asset, and work-demand data—combined with advancements in geospatial modeling tools and techniques—make it possible to quickly and accurately address SDN design challenges.

Introduction

Electric, gas, phone, and cable utilities are being increasingly scrutinized by regulators, customers, media, and shareholders. The result of economic, political, and social forces, this scrutiny is placing more pressure on “wires and pipes” executives to push their companies to higher and higher levels of performance and efficiency. Most recognize that field-force windshield time has a direct effect on productivity, customer/emergency responsiveness, and overall asset services cost. However, many are finding that unnecessary windshield time is not only a product of less-than-optimal scheduling, dispatching, and routing but also outdated "service delivery networks."

A service delivery network (SDN) can be defined as the structure through which services are provided from their source points to demand points. A generalized SDN is shown in Figure 1, where construction, operation, maintenance, and restoration services are provided by field technicians or crews headquartered throughout a network of local service centers. Each service center has responsibility for providing services to customers and distribution facilities within its assigned territory.

Service delivery networks can become outdated if the configuration of service centers, staffing levels, and assigned territories are not altered to reflect changes over time in work demands and

Figure 1 Service Delivery Network (SDN)

concentrations, roads, traffic patterns, and associated costs. Updating an SDN involves answering the following fundamental and strategic questions.

What is the right number of service centers? And, where should they be located?
What should the assigned territories be for each service center?
How many field technicians or crews are needed at each service center?
What effects do service center additions or consolidations have on field service productivity, customer/emergency responsiveness, reliability, and overall network services cost?
Can the same field service level be provided with fewer resources?
What are the estimated costs or savings associated with changing performance by X percent?
From an external service provider perspective, what SDN footprint will be needed to provide the desired level of field service for the work being outsourced?
From an internal service provider perspective, what SDN footprint will be needed for the work not being outsourced? For example, which service centers should be kept and which ones should be sold or leased to external service providers?
If we combine electric and gas asset services, what should the service delivery footprint look like?
After merging with a neighboring utility, will all of the service centers be needed—if not, which ones are kept or consolidated and what are the resulting service center territories and staffing?

Although seemingly very different questions, they relate to the same fundamental issue of optimizing cost and service level trade-offs. Neglecting to effectively answer these questions can result in inefficiencies that needlessly waste resources and undermine the full potential benefits of related improvement initiatives.

This paper looks at the need, and a modern approach, for reconfiguring service delivery networks. It examines how SDNs evolved and could have subsequently become outdated. We will outline the data, geospatial technology, computational tools, and process developments that have enabled us to address challenges related to updating SDN configurations. We will also present the concept, step-by-step process, and application of geospatial modeling techniques for redesigning and optimizing SDNs. The focus of this paper is on field-asset (wires and pipes) service provision. It is therefore primarily directed at leaders of utility distribution companies, and more specifically those—internally or externally—who provide construction, operation, maintenance, and restoration services to related field assets and customers. While this paper highlights issues and developments primarily in the electric and gas utility industries, the stories are similar, and the concepts are applicable to phone, cable, and other field-asset service businesses in general.

An old problem

As the electric and gas utility industries in the United States took shape early in the 20th century, technological advancements and economies of scale significantly decreased costs as production and demand increased. This continued after World War II as electric usage was doubling every nine to 10 years from 1946 to the early 1970s. Residential electric rates (current dollars) dropped from $1.56 per kWh in 1907 to 19 cents in 1947 and to 9 cents by 1967.* Advances in metallurgy, welding techniques, and pipe rolling overcame the barriers of transporting gas from the wellhead to the customer resulting in the construction of thousands of miles of natural gas pipeline. Natural gas consumption grew from 1.6 MMcf in 1931 to 4.2 MMcf in 1947 and 17.4 MMcf in 1967.** Productivity improvements during this period were focused almost entirely on technological advances in the supply and transmission of electricity and natural gas.^{(Hirsh 2001), (Tussing
and Tippee 1995)}

During this time, utilities scrambled to build and expand distribution systems to connect more and more customers and drive early economies of scale in supply. For the first half of the 20th century gas and electric service was, for the most part, localized and concentrated in population centers. Following World War II, traditional population centers were growing and spreading outward to the suburbs as cheaper automobiles provided access to lower-cost housing options.*** Construction, operations, maintenance, and restoration organizations expanded their service delivery networks to keep pace with growth. Developments of SDNs were initially driven more by necessity and compliance than by coordinated planning. Even well-planned SDNs became outdated and inefficient as traffic patterns and geographic concentrations of utility fieldwork changed over time.

In the 1970s, technological progress in supply and transmission failed to mitigate rising fuel, financing, and construction costs causing electricity and natural gas prices to more than double during that period. Customers quickly became disenchanted with the utility system and were demanding better prices and more reliable service.(Hirsh 2001) Congress got involved and passed a series of legislation over the next 20 years to encourage more efficient energy pricing, conservation, and competition in the supply and generation sectors. Distribution inefficiencies that were once overshadowed by productivity and cost improvements in generation and supply were now coming under increased scrutiny as officials gained comfort with more progressive forms of regulation.

Faced with the reality or prospect of rate caps, reductions, or performance-based rates, distribution companies have embarked on numerous initiatives aimed at improving their business performance. Updating the SDN configuration to effectively address current and future work demands has long been a challenge that previously was difficult, if not impossible, to effectively approach. Determining the right number, size, location, staffing, and territory assignment is little more than guesswork without accurate data, computational tools, and an objective process of analysis. Redesign errors or inaction can have costly and lasting consequences. Too many service centers or inappropriate staffing levels cause unnecessary facility and labor costs. Too few service centers or inefficient territory assignments increase windshield time and travel costs, and decrease responsiveness and field-force efficiency. Until recently, these types of problems were approached—if at all—by intuition and “chart, compass, and ruler” techniques. (Ballou)

Enabling Developments

Several key advancements related to computational tools, processes of analysis, and input data have made it possible to address SDN design challenges quickly and accurately. The development of high-level programming languages and powerful mainframe computers in the 1960s and 1970s made possible the broader application of quantitative techniques developed in the 1940s from military and operations research. Industry application of these quantitative techniques—such as linear programming, heuristics, and simulation—for solving resource allocation, distribution, and transportation planning gained value as a decision-making aid.(Shycon and Maffei 1960) Since then, the rapid growth in desktop computing power and the introduction of improved algorithms is allowing organizations to model and solve very large supply-chain problems in a matter of minutes instead of days or weeks.(Bendiner 1998) While most of the early developments focused on optimizing the physical distribution of tangible goods or products, these developments were directly applicable to the delivery or provision of services.

Information technology improvements and integration within utility operations and customer care have generated a wealth of geo-based customer, service center, field-asset, and workdemand data. Geospatial technology advancements in the early 1980s combined the computer display of geographic features, such as points, lines, and polygons, with database management tools for assigning attributes to these features. Likewise, a “location” attribute could be added to traditionally tabular data—such as service centers, customers, and field-assets—and layered with shapes such as roads and assigned territories. The integration of geospatial technology with supply- and service-chain optimization tools created a powerful application for visualizing and solving SDN problems.

A Modern Approach

Advanced modeling tools and techniques that were developed and refined in other competitive, service-oriented industries can be adapted to incorporate the unique operational characteristics of the utility industry. This provides a modern approach for assisting management with answering fundamental questions regarding service delivery network configurations.

Objective

The objective of an SDN reconfiguration effort is to minimize the total long-term related costs (labor, facility, and transportation) for a specified amount of work and desired level of service. Given this objective the key questions to be answered include the following.

What is the optimum number and location of service centers?
What is the optimum assigned territory for each service center?
How many field technicians and crews are then needed at each service center?

In the process of answering these key questions, much can be learned about the effect windshield time has on productivity, customer/emergency responsiveness, reliability, and overall network service costs. Related questions that can be explored include the following.

Can the same field service level be provided with fewer resources?
What are the estimated costs or savings associated with changing performance by X percent?
From an external service provider perspective, what SDN footprint will be needed to provide the desired level of field service for the work being outsourced?
From an internal service provider perspective, what SDN footprint will be needed for the work not being outsourced? For example, which service centers should be kept and which ones should be sold or leased to external service providers?
If we combine electric and gas asset services, what should the service delivery footprint look like?
After merging with a neighboring utility, will all of the service centers be needed? If not, which ones are kept or consolidated, and what are the resulting service center territories and staffing?

Benefits

The benefits of an optimally configured service-delivery network include:

Improved Customer Responsiveness and Service Flexibility
- Decreased average travel time from service centers to job-sites/customers
- Decreased average travel time between job-sites/customers
Increased Operational Efficiency
- Increased “delivery density” – complete more work requests per day
- Minimized travel for given amount of required work
- Optimized number of service centers and required staffing
- Minimized total service-delivery network cost (travel, labor, and facilities)
Minimized Environmental Impact
- Minimized vehicle exhaust/emissions
- Minimized service-delivery network footprint
Enhanced Benefits of Related Initiatives
- Mobile workforce management
- Automated/Optimized scheduling, dispatching, and routing

Base-Case Model

To assist with determining the best long-term configuration, a model of the existing servicedelivery network and corresponding work demands is created using a logistics application and company specific data. The base-case serves as a reference point and validation that the model’s methods and assumptions accurately portray the existing SDN. Validating the base-case scenario helps to ensure that change-case scenarios—which depict network configurations not currently experienced by management— reasonably represent the cost and service levels in practice.

Data Requirements

One complete year of data is typically used--current year data and projections. Data required for developing the model can generally be categorized by its relation to service supply, service demand, and service delivery. Service supply includes data related to the physical supply or “inventory” and location of field technicians or crews. This would include service center locations and cost as well as current staffing levels and related labor cost. Service demand includes data related to the current and projected fieldwork demands expressed geographically. Service delivery includes data related to the road network and vehicles that connect service supply to service demand. (See Figure 2.)

Figure 2 Model Data

Expressing fieldwork demands geographically has become much easier in recent years. Customer information, work management, and geographic information systems provide a wealth of data expressed geographically about different types of work as well as customers and facilities that correlate or tend to drive the level of work (poles, transformers, overhead and underground lines, lights, etc.).

Model Validation

Model validation is the process of comparing model output with actual data and data relationships for the base-case time period. “Flowing” the model is a method of satisfying geographic work demands from available resources across a transportation network. Annual required travel time, onsite time, and travel distance are computed and then compared to actual data, team experience, and expectations. If needed, adjustments are made to input assumptions until the project team is comfortable with the accuracy of the model. (See Figure 3.)

Figure 3 Model Validation

For a model of this nature, it’s important to calculate travel times and distances across a road network, not straight-line distances. Many natural and man-made barriers have a significant effect on SDN design. Digital road networks not only provide an excellent representation of true travel times and distances, but also are reflective of barriers such as mountains and rivers. Change-Case and “What-if” Analysis Once the base-case has been modeled and validated, numerous change-case and what-if scenarios can then be generated and compared to the existing configuration. The types of change-case and what-if scenarios typically evaluated include the following.

No Capital Expenditure Optimization: Optimizing the network configuration subject to existing service center locations and capacities. This is a no-investment strategy where service improvements and cost savings can be realized without an outlay of capital.
Service Center Consolidation: Optimizing the network configuration subject to using fewer service centers.
Service Center Addition: Optimizing the network configuration subject to using additional service centers.
Maximum Opportunity Optimization: Optimizing the network configuration (without regard to capital cost) to identify the network with the lowest possible variable cost. Multiple service center location alternatives are used in addition to the existing ones. These scenarios are useful primarily to identify improvement bounds and additional what-if scenarios
Practical Design/What-if Analysis: Acceptable or attractive network configuration somewhere between no capital optimization and maximum opportunity optimization. Less than optimal from a modeling standpoint, but a better reflection of practical considerations beyond the modeling process.

Scenario Evaluation

Modeling SDN scenarios makes it possible to determine cost and service level trade-offs for various configurations. (See Figure 4.)

Figure 4 Cost – Service Trade-Offs

Comparisons can then be made between the base-case and change-case scenarios to evaluate the cost and benefit of a scenario’s proposed changes. (See figure 5.)

Figure 5 Scenario Evaluation

This information, along with an understanding of the result’s sensitivity to changes in assumptions, can help network services organizations make informed decisions regarding service delivery network configuration.

Which Companies Benefit Msost

As with many types of projects, the benefits realized will vary from company to company. There are many factors that contribute to this, but in general, the companies that realize the most benefits from this type of initiative tend to have the following characteristics.

Travel makes up more than 10-15 percent of productive time
No SDN changes have been made for a long time relative to the rate of change in work demands (including outsourcing changes)
Have merged with or acquired other distribution companies but have not integrated contiguous operations
Larger companies (> 100k customers ~ 25-30 crews)
Primarily contiguous district/service center territories
Network services provider taking on additional O&M responsibilities

We have tried to emphasize the importance of performance improvements for network service providers and identify the many challenges facing their organizations. It is clear that changes in this part of the industry are moving forward rapidly as the pressure to perform at higher levels of efficiency increases. We have outlined a modern approach to address some long-standing challenges and believe that SDN optimization represents a very real, attainable, and major opportunity. Therefore, we encourage executives to look closely at these issues and concepts.

References

Ballou, R., 1992, Business Logistics Management, 3rd Edition.
Bendiner, J., 1998, Understanding Supply Chain Optimization: APICS—The Performance Advantage, January, pp.34-38.
Blumberg, D., 1991, Managing Service as a Strategic Profit Center.
Hirsh, R., 2001, Powering a Generation of Change: Smithsonian Institution, web-site.
Pinkerton, F., 2001, Spreading Out: Electric Perspectives, November-December, pp. 22-32.
Ratliff, H. D. and Nulty, W. G., 1996, Logistics Composite Modeling: Caps Logistics -Technical White Paper Series.
Shycon, H. and Maffei, R., 1960, Simulation—Tool for Better Distribution: Harvard Business Review, November-December, pp. 65-75.
Tussing, A. and Tippee, R., 1995, The Natural Gas Industry: Evolution, Structure, and Economics, 2nd Edition. PennWell Publishing web-site.