Supply process flows are critical to the successful delivery of infrastructure projects and associated key business drivers (growth, reduced costs or lead times, reliability, etc.). Implementing a Project Production Management (PPM) structured approach with suppliers is essential to capture value within the supply network, extending beyond conventional procurement practices. We illustrate this by examining three examples of infrastructure projects.
PPM supports effective partnering with preferred suppliers by providing the means to map, design, control and synchronize process flows between and among the supply network to meet business drivers. Compared to conventional practice, mapping the supply network production system makes risks in the supply chain visible, particularly those beyond first-tier suppliers. It also helps identify ways to reduce cost and better manage business risk through the design of work activities and implementation of control mechanisms to synchronize the process flows across the supply chain production system.
Experience shows that “ways of working” are far more important than the form of contract used to procure discrete pieces of work. Understanding the nature of work activities in the development program (e.g. repeatable standard design, custom design) and the profile of the work activities (e.g. growth in volume, need for flexibility to turn on or off capacity rapidly) is essential to forming the team of suppliers needed to deliver the project. Once in place, PPM supports successful design and control of the planning and execution of work activities during project delivery, while also illuminating areas ripe for joint continuous improvement.
Jennifer Weitzel, PE, MBA, Senior Director, Global Data Center Supply Chain, Microsoft Corporation Industry Council Member, Project Production Institute, firstname.lastname@example.org
Kerry B. Haley, Executive Director, Project Production Institute, email@example.com
This article illustrates different applications of PPM from the point of view of the first author, who was a member of the project team in each case. It is useful to think about the article’s remarks in the context of the PPM framework described in the Overview.
When mapping the Production System implicit in large infrastructure projects, the system typically spans the organizations of several enterprises, including the project owner and the network of suppliers that provide the different inputs (raw materials, parts, capacity, information, etc.) that go into constructing the infrastructure. Figure 1 illustrates a generic example, wherein the production system spans several organizational boundaries, with the dashed lines encompassing the work activities of individual organizations.
Typically, an infrastructure project can have multiple concurrent supply flows. Figure 1, which is an illustration only, shows an example of precast components, structural steel, piping and electrical. Following the Engineering (E), Fabrication (F) and Delivery (D) of the components to the construction site, Installation (I) must take place. The different work activities E, F, D and I are typically performed by different enterprises. For example, the dashed lines surrounding the activities in the precast supply flow and installation activity hypothesize one such division of activity between four different enterprises – one to do the engineering, a second to perform the fabrication, a third to deliver the finished components and a fourth to perform installation. In other words, the production system spans the activities of different enterprises. To properly implement PPM, it is necessary to synchronize and control all work activities in the infrastructure project across the entire production system, spanning the organizational boundaries of the enterprises working throughout the supply network.
The three infrastructure projects described within this article illustrate the efficacy of the PPM approach. Within the context of the PPM framework, as seen in Figure 2 below, the examples demonstrate how the levers of capacity and inventory are used to manage variability. Oftentimes, these levers lie within the purview of supplier partners rather than the project owners. In some cases, process redesign also mitigates variability.
The three examples demonstrate different aspects of production system mapping, design and control, all within the context of the above framework. The first example, Heathrow Terminal 5, is about the design and control of supply flows for civil works. The challenge in this case was about getting the right materials and information to the right place at the right time by effectively synchronizing supply with construction site demand, based on a set of given constraints such as site access, work hour requirements, limited laydown space at the construction site, etc. In terms of Figure 2, the example illustrates the value of mapping the production process (transformation, flow and stocks), identifying the impact of constraints on the supply flows and implementing mechanisms to control work-in-process (WIP) and minimize variability (while it is impossible to eliminate all variability, it can be minimized).
The second example on supply flows for the construction of retail outlets illustrates the value of mapping production flows to assess supply chain complexity so that opportunities to reduce cost are identified while ensuring responsiveness to the variability in demand. Within the context of Figure 2, the case entails knowledge of variability and process design to thoroughly understand where the cost reduction opportunities are available through the appropriate negotiation and partnering of suppliers, possibly including the reconfiguration of supply processes for one or multiple suppliers.
The third example of an owner building data centers demonstrates how value is delivered to the end users (data center users) by ensuring that the supply of skids (engineered-to-order or ETO items) did not unduly constrain construction. At the same time, opportunities were exposed to reduce lead times and supply costs through comprehensive process design of the work activities and identifying where unnecessary resources were (time, capacity and inventory) throughout the project production system.
Example 1: Heathrow Terminal 5
Heathrow Terminal 5 was a massive infrastructure project that has been written about extensively due to its many groundbreaking and innovative aspects, and most of the literature [2 – 6] references the Project Production Management aspects. It was a very large and complex airport transportation hub project, with a total project cost of £7B comprising 16 discrete projects and 147 subprojects. These included: HATCT, HEX, the London Underground Piccadilly Line Extension, Terminal 5, A, B, C Concourses, the baggage system, new control tower, civil work onsite and external improvements. Beyond these projects, the T5 Project also required the rerouting of two rivers and integration with the M25, which was then Europe’s busiest motorway. Adding to this inherent complexity and sheer magnitude, the civil phase of the program alone was executed based on 80 concurrent subprojects, requiring 3,500 craft workers supported by 2,500 engineers and administrative personnel. The physical scale and scope of this project were enormous.
The project delivery team agreed on the following guiding principles to design and control supply flows, with demand coming from the site based upon given constraints. To achieve this, the team made the following commitment:
“The right information, the right materials, the right labor and the right equipment, in the right quantity, would be delivered to the right place at the right time, every time.”
To say there were some unique challenges faced by this project would be an understatement. The ingress and egress to the construction site were highly restricted: literally one way in and one way out for all traffic. This was due to the need to maintain ongoing operations for the rest of the airport, one of the busiest in the world. There were also specific constraints imposed by the surrounding community, such as lorry deliveries could occur only during certain hours on certain days to minimize disruption. This placed severe limitations for scheduling deliveries of supplies and materials to the construction site. In addition, the available space to store inventory on the site was limited to the amount of materials the construction site could “consume” over the next 24-hour period, period. Given it was a public infrastructure project, beyond deliveries to site, it was necessary to organize the work activities in construction and installation to minimize overall impact on the surrounding community. Clearly, health & safety on and offsite were paramount. This was an asset that had to operate flawlessly from Day 1, which entailed very careful and proactive management of risks to ensure that outcome was achieved.
Following the example of BP Andrews [7 – 9], T5 implemented a form of collaborative construction agreements to align the incentives of all participants toward overall targets for infrastructure delivery. The project team also adopted a single construction-modeling environment to visualize the overall infrastructure and to understand where, for example, issues in assembly and installation might occur, and then to systematically design and schedule the work activities to avoid such issues.
Given its complexity, a comprehensive process flow mapping study was required. The application of Operations Science principles to identify the points of variability across the entire production system map and simulate the system behavior, identified several hurdles and assisted in quantifying opportunities for optimization within distinct supply flows feeding the construction activities. For instance, the case of the expanded polystyrene panels supply flow mapping, configuration and control, is dealt with in detail in [5, Case Study 2, Figures 5 and 6]. In consideration of given constraints, the mapping process soon identified elements of the project that could not be assembled or manufactured onsite. All of this entailed the design and definition of 1-2 day production packages that could be held onsite, which necessitated implementation of a CONWIP (constant work-in-process) control protocol to effectively limit and control WIP, together with offsite assembly and logistics centers . It required the leveling of site demand with the supply network, which allowed for WIP reduction including fewer deliveries and only when site space was available to hold the production packages – for example, controlling the production and delivery of rebar assemblies . Process flow mapping and identifying simple cost reduction opportunities, e.g., storage, handling and freight costs, etc., allowed the project team to put control mechanisms (to enable CONWIP and Pull protocols) in place that assured the smooth flow of work completion, respecting the constraints on logistics within the project.
When delivering complex projects through the use of PPM, “ways of working” are created that foster collaboration with suppliers. Employing techniques like Process Flow Mapping and Production Control are key to understanding opportunities and risks while driving reliability into the project delivery process without compromising health, safety and quality / operability.
Example 2: Retail Infrastructure Development
The second example is that of the retail infrastructure program of Tesco PLC, a United Kingdom retail supermarket chain. During the years observed, the program in question involved capital investment ranging between £500M-1B per annum to deploy new sales square footage for Tesco. Business revenue growth was in fact partially driven by the rate at which deployments (e.g., new retail sales areas) came online. Through the use of Process Flow Diagrams, the team was able to assess the complexity of the supply network, its capability to respond to Tesco’s dynamically changing needs and expose opportunities for cost reduction.
The program was one in which most project deliveries were repeatable and based on a prototype design. Also, fast and nearly instant feedback on ROI was forthcoming, generally received soon after a store opening. The annual development program consisted of a high volume of projects, typically over 100 per year. The project portfolio consisted of a handful of Flagship projects, or pilots, followed by a series of new stores or extensions as well as refreshes of older assets. Business drivers in project design included: cost reduction, agility and responsiveness to changing customer needs, and fast project delivery tightly coupled with project initiation.
A challenge in executing the program is exemplified by a quote from an employee: “Our suppliers have become so embedded in how we work that they have lost the ability to innovate and create change.” This was directly at odds with the objective to reduce costs and implement the changes necessary to rapidly respond to changing customer needs.
In order to achieve program objectives, it was necessary to work with suppliers to map out the full cycle of activities over the life of a project and across the program. This included identifying potential cost reductions across land acquisition, product design, materials specifications and delivery methods, among others. Typically, the process consisted of the following steps:
- Select critical supply flows based on impact on delivery and cost
- Plan visit to supplier / fabricator
- Map current state of production process (transformation, flow, stocks, control mechanisms, etc.)
- Analyze and identify opportunities for optimization
- Implement actions to improve working with suppliers
To better illustrate the application of this process, we describe two critical supply flows: the production and delivery of terrazzo from raw material to installed product, and the production and delivery of multiple components of the refrigeration system, a key aspect of supermarket stores.
Terrazzo Supply Flow
Terrazzo is a composite material, either poured in place, or precast, which is used for floor and wall treatments. The Terrazzo supply chain was one such example of a critical supply flow for Tesco. The supplier produced a mix of made-to-stock and made-to-order components, at that time with an annual expenditure of £18M per year (for materials and installation) using a single installer and fabricator, affecting 119,000 square meters per year. It was a critical supply flow for both parties, given that Tesco represented approximately 90% of the supplier’s business.
Rather than place phone calls with suppliers, the practice of visiting suppliers and walking the site to observe the current state of the production system at work generally afforded opportunities to identify bottlenecks and areas where additional capacity might improve the flow of work through the system.
For example, the photographs in Figure 4 show different places where inventory or WIP was building up in the system, and raised questions about whether batch sizes at different points in the terrazzo production were appropriately sized for the needs of Tesco and its business / program objectives, or whether they were in fact adding cost and slowing down production.
The map of the current state of the terrazzo production system (seen in Figure 5) was the result of working with all the suppliers to characterize the supply flow from beginning to end, paying attention to stock / queue levels in front of each work activity, denoted by the triangles. By analyzing the system, several significant improvements were identified and implemented, illustrating some general principles of production system optimization as described below:
Determine basis of optimization: As efforts to understand the current state of a production system already designed (like the case of Terrazzo) are undertaken, it is critical to determine the basis by which opportunities for optimization will be identified. For instance, if we are trying to optimize for time (i.e. compress schedule), using time buffers to deal with variability as well as excessive WIP (time is needed to build WIP) will be counterproductive to the objective.
Trigger work at the most cost-effective point in the delivery process. An example is addressing environmental contamination remediation at the land assembly stage sooner rather than later. One example was spraying the Japanese knotweed years in advance of needing the store, at a cost of £40K, versus removing the knotweed and replacing it with engineered fill a year before the store was required, with a comparative cost of multiple millions of pounds sterling.
Go to the work. Visible and tangible clues of capacity constraints and how a supplier mitigates these constraints (Terrazzo inventory stockpiling) are critical to understand.
Ensure there is transparency of how you’re buying and any impact or risk to supplier. It was key to recognize that Tesco represented 90% of Terrazzo Suppliers’ annual production, as well as the risk of potential disruption of supply in the future.
Ensure there is transparency of how your specifications impact and/or add risk to a supplier. Custom specifications limit flexibility and nimbleness to change, so it was important to collaborate with suppliers to lower complexity of specifications.
Refrigeration Supply Chain
A second example of a critical supply flow in the Tesco retail infrastructure development plan was the refrigeration system supply chain. This was also comprised primarily of made-to-order (MTO) and engineered-to-order (ETO) components, an expenditure of around £80M per year split between two suppliers. One supplier had 80% of the business and the other supplier had 20%. Tesco represented about 50% of one supplier’s business and 41% of the other supplier’s business.
In pursuit of the most favorable technologies from a sustainability perspective, Tesco assessed its refrigeration solutions and viewed this as an opportunity to strategically re-evaluate its supply base, both in terms of its costs and its capabilities. As illustrated in Figure 6, simply walking around the site helped suppliers and Tesco map the entire production system, as depicted in Figure 7. A key question was assessing the supply chain’s responsiveness to changes in demand. This question was provoked by the long cycle times embedded in the various steps within the existing supply chain, implying that product specifications needed to be locked in with long lead times in advance.
As with the Terrazzo supply flow, inspecting the different sites along the entire refrigeration supply flow culminated in the process map illustrated in Figure 7, distinguishing which suppliers made-to-stock (MTS) and which ones made MTO or ETO. The cycle times for each step in the supply chain were also measured. Given the cycle time between starting an order and delivering a finished refrigeration unit, it was critical understand how quickly the entire supply flow could respond to the changes in demand.
Mapping the supply flow in collaboration with the supplier helped identify opportunities to reduce cost and cycle time, thus improving the responsiveness of the supply chain and resulting in roughly £120M of value that was ultimately captured across the entire infrastructure expansion program.
Example 3: Technology Infrastructure Program
The third example is from a technology infrastructure program, specifically focused on data centers. The project owner had two data center products, one for electrical power supplies working at 50 Hz frequency and the second for electrical power supplies working at 60 Hz frequency. The data center products came in two different types – a standard repeatable pod design or an engineer- and fabricate-to-order custom solution. The two data center products were addressed with two separate production lines, with different project lead times, and at different price points (the basic unit of measurement is $/MW IT Load).
Eighty percent of projects in the program were delivered globally, using standard designs combined with minor adjustments to meet local site conditions and planning requirements. The design, supply chain and delivery model permitted the phased delivery of 10k square foot tranches of 1-2MW IT load capacity within the data centers. Rapid growth in the industry drove opportunities to capture market share with a swift deployment model delivering small amounts of data center capacity in multiple core markets (from a production perspective, a strategy that reduces batch sizes to enable faster delivery).
The project owner used traditional forms of contract with general contractors (AIA forms in US, JCT in the UK, etc.). For equipment with 6-month lead times, the project owner held, and replenished as consumed, a conservative volume of inventory. For other long lead-time equipment (i.e. 12-20 wk. lead times) vendor-managed-inventory (VMI) agreements were established with preferred suppliers. Figure 8 shows examples of the types of equipment that were followed through the supply chain, while Figure 9 illustrates the same steps taken to map the production system and optimize the supply flows.
Figures 10 and 11 illustrate the production system process maps that were produced. The steps taken to control the supply flows are described below.
A few general principles emerged from the analysis of the current state process map of the production system. These primarily involved understanding the production flow and where things go wrong through a basic assessment of construction schedules and areas prone to variability. Some of the most important principles are described below, with examples on the changes that were made to the production system.
- Tackle sources of variability, but prioritize lowest effort and highest impact. These efforts were the first to demonstrate success. Basic analysis of the construction schedules illustrated that approvals processes and quality issues related to electrical installation, which occurred during peak production onsite and during “dirty works,” caused a lot of variability in completing IST (integrated system testing) prior to project completion.
- Determine strategy for offsite assembly. Understand and measure the benefits of deploying modular or offsite assembly methods. Electrical skids were pursued to improve quality, improve reliability of delivery and simplify processes onsite. These changes yielded better product quality and improved reliability of delivery. However, it is important to point out these were NOT cheaper. Additionally, as these skids were ETOs units (they go to a unique place in a data center and are not interchangeable), synchronization of supply and demand is key (inventories cannot be used to synchronize as in made-to-stock supply).
- Processes and standardized designs are never as simple as originally intended, and often have unintended consequences. Process Flow Mapping frequently unveils some surprising areas of variability, combined with opportunities to simplify and improve reliability. The owner had standardized on a repeatable design for electrical skids. However, there were different seismic requirements that introduced variability in steelwork manufacturing, two different UPS suppliers and two different Switchgear suppliers. The electrical skids weren’t actually standardized, so simplification was imposed by matching UPS and Switchgear combinations and standard skid heights to accommodate requirements. These changes resulted in significantly reduced variability.
- Develop and maintain transparency into 2nd & 3rd tiers of the supply chain. This offered the project owner opportunities to understand risk that the owner may not have otherwise been aware of. Some examples include:
- Electrical skids: By purchasing small inventories of switchgear and UPS, and holding it at their facilities, both the number of steps in assembly and lead times were reduced. This simplified the process and shortened cycle times to produce skids.
- Equipment Manufacturer Quality Control issues: A root cause analysis exercise highlighted that the two manufacturers shared the same component (transformer) manufacturer in Asia. Having transparency to this allowed identification of risks and, should quality issues occur, the ability to manage across suppliers.
- Use the right tools to unlock opportunities. Vendor Managed Inventory was used to shorten lead times for equipment from the traditional 12-16wks to 5wk lead times.
- Right size inventory of critical components, but not necessarily finished goods. No production system including those of projects can operate without some level of inventory (stock and WIP). Skinny inventory levels were the goal, with “just right sized” versus “just in time” to deliver shortened lead times for infrastructure projects while not risk running out of stock.
- Consistent communication with partners through weekly production planning. This enabled better control of how capacity gets allocated including cycle time reduction opportunities and the associated rapid responses to changing business needs.
- Annual goal sharing, technical roadmap sharing and reviewing (two way) performance. This created opportunities for success for the mission ahead. Frequent forecasting and sharing of potential projects or business drivers enabled deeper understanding and better preparation to respond to peaks or troughs in production.
As a result of this exercise, several significant results were achieved, summarized below and in Table 12:
Reduced Delivery Lead Times
Inventory held in primary markets with Generator Manufacturers was pulled to manufacturer enclosures. For all equipment with 12-16wk lead times, the project owner entered into Vendor Managed Inventory agreements with the primary supplier in each category, thereby reducing product lead times to 5wks.
Improved Reliability & Quality
PPM helped to refine the delivery and reliability of offsite assembly of skids for electrical plant rooms. Process Flow Mapping and ongoing production control reduced the overall cost of this assembly process and provided the team with visibility to opportunities for further cost reductions.
|Process Design (Number of Steps)||252||97||Minimize steps outside standard processes and explore unknown steps|
|Supply Reliability||0-38%||98%||Sustain reliability levels through control|
|Cycle Time (PO to Install)||65-220 days||Target – 71 days Actual – 49 days||Further reduce variability to reduce overall cycle time|
|Cost||100%||64%||Decrease time for onsite installation to further reduce cost.|
Implementing a PPM-structured approach with suppliers is essential to unlocking and capturing value within the supply network. The use of Process Flow Diagrams to assess supply networks has the added benefit of providing transparency for a project owner beyond their first-tier suppliers. This also establishes a platform for the identification of optimization opportunities, allowing them to identify risks across the entire supply chain and identify the participant(s) best positioned to manage the risk. PPM also allows the owner to strategically establish areas where collaboration with suppliers is key to 1) effectively reduce variability, 2) adjust process design as deemed appropriate and 3) quantify and place inventory or capacity buffers to mitigate remaining variability.
Mapping the entire production system and working with suppliers to implement improvement opportunities throughout the entire supply network provides benefits to all participants in the project execution and delivery process, including the project owner and the supplier network. Experience shows that “ways of working” are far more important than the form of contract used to procure discrete pieces of work.
Once in place, PPM underpins successful design and control of the planning and execution of work activities during project delivery, while illuminating areas ripe for joint continuous improvement.
- J. Weitzel, “Project Supply Chain Optimization”, presented at the Project Production Institute Symposium, 30 November 2016.
- T. Brady and A. Davies, “Managing Structural and Dynamic Complexity: A Tale of Two Projects”, Project Management Journal, Vol. 45, No. 4, 2014, pp. 21 – 38.
- R. Arbulu, G. Ballard and N. Harper, “Kanban in Construction”, Proc. IGLC-11, 2003.
- R. Arbulu and G. Ballard, “Lean Supply Systems in Construction”, Proc. IGLC-12, 2004.
- R. Arbulu and G. Ballard “Linking Production-Level Workflow with Materials Supply”, Proc. IGLC-13, 2005 pp. 199 – 206.
- R. Arbulu, “Application of Pull and CONWIP in Construction Production Systems”, Proc. 14th Annual Conference of the International Group for Lean Construction, July 2006, pp. 215 – 226.
- J. Pratt, T. Priest and C. Castenada, Offshore Pioneers: Brown & Root and the History of Offshore Oil and Gas, Gulf Publishing Company, May 2011.
- A. Bakshi, “Alliance changes economics of BP Andrews Field Development”, Offshore Magazine, Vol. 56, No. 1, Jan 1, 1995.
- M. Sakal, “Project Alliancing: A Relational Contracting Mechanism for Dynamic Projects”, Lean Construction Journal, Vol. 2, No. 1, 2005, pp 67 – 79.
About the Authors
Jennifer Weitzel, PE, MBA
In the role of Senior Director of Global Supply Chain at Microsoft, Jennifer Weitzel is responsible for overseeing all facets of the Global Supply Chain in support of the Data Center delivery team. Prior to joining Microsoft, she spent 7 years at Digital Realty serving in roles ranging from Director of International Construction to most recently, SVP Global Supply Chain. She led a team who partnered with preferred suppliers to meet business objectives: delivering new Data Center assets, maintaining the existing portfolio of assets and leading power procurement. Weitzel’s previous experience includes delivering retail stores for global retailers in roles such as Director of Construction Delivery at Tesco, plc., Consulting Engineer for Walmart and construction of a major project at Heathrow Terminal 5.
Weitzel is a Canadian Professional Engineer, holds a B.Sc. in Applied Science and Engineering from Queen’s University, and an MBA from London Business School where she served on the Governing Board and as President of the Student Association. In addition, she studied Leadership in Supply Chains at Stanford University.
Kerry B. Haley
Kerry Haley brings extensive managerial and operational experience including application of Project Production Management (PPM) principles, tools and methods across the global energy, telecommunications and civil infrastructure sectors.
In her role as Executive Director, she is responsible for global outreach to promote awareness of the Institute’s purpose engaging influential members of academia, industry, science and technology.
Previously, as Director of global infrastructure at Better Place, Inc., she was responsible for developing the methodology and implementation plans to design, deploy and operate EV charging infrastructure at scale worldwide. Prior to Better Place, she held executive positions in the global wireless telecommunications sector for a mesh network systems supplier and led the creation of the wireless division of Frontier Communication (NYSE: FTR).
Haley earned her B.S. in International Finance from Santa Clara University.