Human resource allocation in a multi-project R&D environment

International Journal of Project Management Vol. 17, No. 3, pp. 181±188, 1999 # 1998 Elsevier Science Ltd and IPMA. All rights reserved Printed in Great Britain 0263-7863/98 $19.00 + 0.00 PII: S0263-7863(98)00026-X Human resource allocation in a multi-project R&D environment Resource capacity allocation and project portfolio planning in practice MHA Hendriks, B Voeten and L Kroep Pasteurlaan 67, 5644 JB, Eindhoven The Netherlands In their strive for more e ective usage of scarce resources R&D organisations tend to seek the solution in forecasting in more spliced tasks and by allocating the most skilled person for such a small task. To get this done the aid of sophisticated project management software is evaluated on the possibilities of coupling these ne day-to-day tasks to the project portfolio plan as presented in the business plan. During the study by the authors of this article on how to optimise an existing resource allocation process in a large R&D organisation it was found that a breakthrough in the allocation method could be made by introducing two indicators: namely, the project scatter factor and the resource dedication pro le. By giving attention to these indicators the resource allocation process was signi cantly simpli ed and, the most important, better project and business results were made. # 1998 Elsevier Science Ltd and IPMA. All rights reserved Keywords: project scatter factor, resource dedication pro le, human resources, R&D organisation, project portfolio, business plan, project planning, project management, project management software Introduction Ultra short time-to-market and low costs of projects require e ective and e cient project oriented organisations. As progress in projects is not always as what is foreseen and because new projects need to be started up all of a sudden during the year due to fast changing results of, e.g. competitors an adequate human resource allocation process is essential. Allocating people to projects in multi-project environments is dif- cult and often faces lots of problems. Important in this process is the coupling of day-to-day planning for each individual person to the long-term business plan. Our study on how to optimise an existing resource allocation process in a large R&D organisation pointed out that ve elements are vital in the set up of an adequate resource allocation process. Central element is the so-called medium-term-resource-allocation. Main output of the medium-term-resourceallocation is a `rough-cut-capacity-planning' that shows what projects will be active for the coming months including a rough allocation of the resources for each project. Multi-project environment In research and development organisations human knowledge is the most important and most scarce resource. Allocating the right human resources to a project is vital. The more projects that are involved and the more speci c knowledge that is needed in every project, the more important, but also the more di cult, is the allocation process. In many R&D organisations the matrix structure is commonly used (Figure 1). According to literature as well as in practice the methods used for resource allocation in such matrix structure are various. Most organisations are looking for an MRP-like system based upon Bills of Materials where projects are treated in the same way as goods that can be ordered with a certain delivery time. Therefore, everyone starts with looking for integrating their used project planning system for single-project situations on a day-to-day basis. They want to consolidate the data of all current and future projects with one push on a button while on the other hand the horizon is extended from short term to medium term. Because of the great 181

Figure 1 Multi-project situation in a matrix organisation uncertainty in the planning of projects in an R&D environment the detailed task planning is only useful when updated frequently. When not up-to-date, planning is useless. Therefore all activities need to be entered into the system on a detailed level otherwise the outcome is too unreliable. Also new projects that maybe will be activated in the portfolio of projects have to be entered in such a detailed way. Resource allocation methods During the last years several authors wrote on the subject of resource-allocation for projects. These articles focused mainly on a method for the short-termresource-allocation: a model for day-to-day planning and priorities is given. The articles present a theoretical review and implementation issues are not mentioned. In Pillai 1 a resource-allocation for a longer period than short term level is mentioned but the short-term-resource-allocation is missing. In 1994 a `rough-cut-project-and-portfolio-planning' was published by Platje et al. 2 In this method the planning and control cycles for single projects and the portfolio of projects is made (see Figure 2). Rough-cut-project-and-portfolio-planning In 1994 we implemented the concept of the `rough-cutproject-and-portfolio-planning' in a young R&D organisation that had grown very fast towards 200 professionals in various disciplines working in 80 di erent projects. The allocation process was designed in such a way that every quarter an inventory was made on a proposed project portfolio by using simple `Resource- Claim' and `Resource-O er' spreadsheets in the matrix organisation (see Figures 3 and 4). This provided a quick and global overview of the requested project portfolio. Each claim and each o er was made clear by a percentage of workmen needed for the next quarter. The claims by the project leaders needed to be discussed with the group leaders. By doing this the `rough-cut-project-and-portfolio-planning' was a tool to have the project leaders contact the group leaders in a structured way. Also management was forced to decide on what projects were placed in the portfolio and a decision on the allocation of the resources. Without such a quarterly `rough-cut-capacity-planning' the decisions on the portfolio were made too late and due to the pressure of all days often no time was left for a rough resource discussion between project leaders and group leaders. Five elements In 1996 a study took place by the authors of this article on how this `rough-cut-capacity-planning' method could be further improved. A great number of small projects were not explicitly looked at so big projects got almost all the resources. Furthermore, the link to the day-to-day resource allocation, answering whether resource X needed to work on project A today and on project B tomorrow afternoon, was very weak. In our study we found that ve elements are vital for human resource allocation in multi-project situations. These ve elements are:. long-term-resource-allocation;. medium-term-resource-allocation;. short-term-resource-allocation;. links;. feedback. The relationship between these ve elements is shown in Figure 5. By using these elements in a correct way this allocation method gives an organisation a exible day-to-day planning based on and in line with the business plan. Long-term-resource-allocation Seeing that a certain discipline within an R&D organisation is understa ed and the moment that such a discipline is sta ed appropriately takes several months, at least. Therefore a long-term plan of the needed resources is required. A long-term plan is based upon the business plan that speci es what the needs are for each discipline, for at least the coming year, and therefore is translated into yearly budgets for the departments and groups. Distribution of the resource budgets must give a rough indication of the needed e orts of a discipline. A discipline where e orts will decrease during the coming years must get a decreasing budget, disciplines where e orts will increase must get more resources and therefore an increasing budget. Portfolio planning process in multi-project organ- Figure 2 isation 182 Medium-term-resource-allocation A periodical review of the project portfolio is needed. Using day-to-day planning for this purpose seems to

Figure 3 An example of the collected `resource claims' be undesirable. This kind of planning often is very unstable. In contrast, the project portfolio must be more or less stable. Planning the portfolio once a year can be e ective but looking at development organisations in practice this review time is too long. Changes in the portfolio within a year are inevitable. Therefore a new level of resource allocation must be set up to determine the project portfolio. The main input for this mediumterm-resource-allocation is the long-term-resource-allocation. The output must be in line with the short-termresource-allocation (Table 1). The main characteristics of the medium-termresource-allocation are:. The contents of the project portfolio. The mediumterm-resource-allocation can be a good tool to provide a link between the budget and day-to-day planning when the right projects are chosen. Therefore, the rst result of the medium term must be the contents of the portfolio;. Although the budgets and project portfolio generate a realistic portfolio there is always a possibility of resource overload. To give the project leaders and group leaders a hand to solve these problems the medium-term-resource-allocation has to give decision rules. The main goal of these rules is to make clear which task has to be executed rst in case of resource con icts;. When the portfolio is determined and the main decision rules are made a rough-cut-capacity-planning must be set up where the resources over the projects 183

Figure 4 Flow chart medium-term-resource-allocation are assigned roughly. This is important input for the short-term-resource-allocation;. The rough-cut-capacity-planning has to be an agreement between the project leaders and resource leaders. Mainly, the claims for resources by the project leaders together exceed the number of resources. The project leaders know the resource planning and therefore they can estimate the project delay. The project delay can be communicated with all people involved and eventually a new planning can be made. weeks. Now almost all deviations can be treated by group leaders in close harmony with the project leaders. Interference by management can in this way be very limited. Links The long, medium and short-term allocation process have their own goals but together they must be linked to provide the organisation with the needed results for doing business. These links must give the information needed to make the right decisions. Short-term-resource-allocation Using the rough-cut-capacity-planning and the decision rules as main input, the short-term-resource-allocation must be the main input for the day-to-day planning of the individual resources for the coming Evaluation The links give input to make the right decisions. This input can be made better by evaluating the input vs the real e ort, when the evaluation must be used for the long and medium-term-resource-allocation. The allocation process can be made better by the feedback. Figure 5 Links between the various resource allocation processes 184

Table 1 Three resource allocation processes with their speci c goals Resource allocation process Purpose Output Frequency Horizon Long term Medium term Short term needed capabilities for accomplishing the Business Plan rough cut capacity planning for the project portfolio operational day-to-day assignment of people. Department plan, budget per capability. portfolio check, which projects must be executed. decision rules for group leaders. analyses of the e ects on the milestones of the projects (changes in targets). agreed rough allocation as input for the short term resource allocation. assignment of tasks to persons, within the medium term resource allocation assignment yearly quarterly bi-weekly 5 years 21 year 26 weeks Growth model The ve elements used in this method can be implemented in almost every organisation. How, and to what level, the ve elements can be implemented is situation-dependent. In Table 2 a model of how to grow to a mature level of resource allocation is presented. R&D environment Allocation of workmen in a multi-project environment is an issue for lots of organisations such as software houses, R&D-organisations, construction and engineering companies. This article especially ts in with R&D-organisations where multiple projects are running concurrently. Speci c for R&D is:. Project results and project timing are very uncertain because of the unique characteristics of each project based upon a high degree of innovation;. Human resources are the main and scarcest resource in R&D projects. Knowledge is scarce. Therefore almost everybody provides his or her small speci c contribution to every project;. The progress of each project strongly depends on the state-of-the-art building blocks that need to be `invented' by scientists and engineers. Human ingenuity is very much dependent on the motivation and involvement of each individual engineer. These three characteristics make the process of resource allocation di cult. The allocation method needs to be exible because of a fast changing project portfolio. Project scatter factor In an R&D environment the allocation of specialists to one project may be di cult because of the speci c knowledge most of the people have. The number of sta needed to ful l a one year task is what we called the `project scatter factor' (see Figure 6). A higher `project scatter factor' means more people needed for one task and therefore the project team will be bigger and the devotion and e ciency of the work done within each project by the experts decreases. In Figure 7 the e ect of the `project scatter factor' is shown. Devotion of the sta and also the e ciency is shown. As stated in Berenschot the `project scatter factor' has great in uence on the project progress. ``An acceleration in the product development process is only possible in a good project organisation with empowered project leaders and by allocating people fulltime (or almost full-time) and working as a team under control of the project leader'' (Bureau Berenschot by order of the Dutch Ministry of Economic A airs, Report on Technology Reconnoitre, 1995) Therefore a `project scatter factor' of 1 is most desired. Hammer and Champy 3 also stress this by saying a project team has to be a small multi-functional team: the `project scatter factor' should be as low as possible. Resource dedication pro le To realise a small project team the resources must have as what we call the right `resource dedication pro le'. In most R&D organisations the resources have their own specialisation. This knowledge is concentrated on a small piece of work to be done. This greatly in uences the `project scatter factor'. If many Table 2 Element Growth possibilities of the ve vital elements for the resource allocation method Growth possibility Long-term resource allocation not done 4 assignment of budgets related to business plan Medium-term resource allocation not done 4 set up a project portfolio related to business plan, assignment of tasks to persons/disciplines and decision rules Short-term resource allocation ad-hoc 4 resource leaders assign persons according to medium term allocation Links not available 4 synchronisation of the allocation processes, execution of the allocations on the right level, data exchange between the allocation processes Feedback not available 4 feedback of the tasks, the projects and the portfolio in all three allocation processes 185

Figure 6 Project-scatter-factor people are needed in one project on a part-time basis, a high `project scatter factor' will be the result and the project e ciency will be low. The problems faced in a R&D situation can be solved by changing the `resource dedication pro le' (see Figure 8). Specialists should be transferred to Figure 7 E ects of the project-scatter-factor 186

either all-round project members or to experts or to service employees. All-round project members The rst-group of resources must be at the heart of the project teams. These people must have general knowledge of the whole project and must be exible within the project. So they are the rst step towards an e cient and e ective well-organised project. The main de ections within the project must be caught by the project members themselves. Above this, allocating people to just one project makes the allocation process far easier. There are extreme situations and problems which cannot be solved by the project members. For these types of problems the knowledge of experts with extensive knowledge of a certain discipline will have to be used. Experts A small group of people in an R&D organisation must have very speci c knowledge that is used within the projects. Experts can and must be used in situations of problem-solving or at project reviews. The project members often don't have the deep knowledge to tackle severe problems. Experts have to provide this knowledge. They give new and extra information to the project. The experts will be needed in all the projects in the portfolio several times. When they are needed is hard to predict. They cannot be allocated to just one project. They are the coach and main source of information for the project leaders, project members and group leaders. The experts have a special place in the allocation process. It isn't possible to predict their e orts in each individual project for a long period (more than 1 or 2 weeks). Therefore, we give the advice to predict the number of experts needed in the long-term-resource-allocation and give them freedom by not allocating them in the medium-term-resource-allocation. In the shortterm-resource-allocation they must plan their own work in co-operation with their group leader. Service employees Although the projects in the portfolio all have unique characteristics there are certain activities which must be executed in every project. These are routine-like activities which can be done by several people of a Figure 8 Resource dedication pro le certain discipline. When this activity is routine-like, independent of the chosen resource and is constant of quality this activity can be de ned as a service. The projects then do not have to worry about these activities. These services can be asked for, internally or even externally, when they are needed. The services can be planned independent of people. In the medium-termresource-allocation only a check of the availability versus the need of each kind of service is necessary. By planning services instead of people the complexity of the medium-term-resource-allocation process decreases and the work can be organised more e cient. Consequences Allocation of the resources according to the new `resource dedication pro les' will substantially in uence the performance of an organisation. But changing the resources isn't that easy. Therefore the whole organisation must understand the possibilities of the changes. First, discipline leaders must see the need of changing specialists into project members, experts and service employees. Only a few people are needed as experts. On the other hand, the organisation must seek for the services. De ning more services gives better controlled projects and the resource allocation process is easier to execute. This will greatly in uence the career paths of the employees. Experiences After having formulated the resource allocation problems we came to the conclusion that the only viable way is to go in the direction of multiple mono-project situations despite the initial arguments of discipline leaders that it would be impossible to get all-rounders. The e orts to get e ciency between projects is far less e ective than the e ciency that can be gained within each project itself. Until this study, the ve elements of resource allocation were not used explicitly. Now they are placed in the right business process. The long-term-resource-allocation is placed in the strategic business process. Each year this resource allocation process is walked through. Direct linked to this process is the medium-termresource-allocation. This process is executed each quarter. The day-to-day planning is used within the development and the production process that needs several `development resources' to optimise the production process. By making these three processes explicit in the business process map of the organisation it was now possible to explain to the organisation what the goals were of these processes. The involved persons now better understand the overall need of a resource allocation process. In the past two quarters we adapted the mediumterm-resource-allocation process according to our new insights using the `project scatter factor' and the `resource dedication pro le'. After the introduction of these two indicators the organisation decided to stop with their strive to nd a computer-button mediumterm-resource-allocation. Maximum e ort of the ve elements, by linking them and by optimising each element, does not give as much pro t as the usage of the `project scatter factor' and the `resource dedication pro le'. 187

The allocation process is now less complex and the throughput time is shorter. Computer planning software, as linked between the medium-term-resource-allocation and the short-term-resource-allocation, is not necessary even in an organisation with more than 200 people involved on 80 di erent projects. Medium term planning is easy to handle and exible with a spreadsheet and the day-to-day planning can be executed by each individual project leader and the group leader. Ir. Martien H.A. Hendriks graduated as an electrical engineer at Eindhoven University of Technology in The Netherlands in 1979. He has been a manager of a service engineering department within Philips Electronics. Since 1991 he is a Project Management/Business Engineer. Currently he is a Business Engineering and Information Technology Manager at Philips Optoelectronics in Eindhoven. Keep it simple Many organisations have a day-to-day planning method where all detailed tasks are allocated with usage of planning software to individuals. In these organisations a medium-term-resource-allocation is not used. But to couple the day-to-day planning to the strategic business plan this medium-term-resource-allocation is absolutely necessary. When using the `project scatter factor' and the `resource dedication pro le' a medium-term-resource-allocation process can be implemented very easy. Moreover, the project results will increase rapidly. References 1. Pillai, A. S. and Tiwari, A. K., Enhanced PERT for programme analysis, control and evaluation: PACE. International Journal of Project Management, 1995, 13(1), 24±29. 2. Platje, A., Seidel, H. and Wadman, S., Project and portfolio planning cycle, project-based management for multiproject challenge. Internal Journal of Project Management, 1994, 12(2), 100±106. 3. Hammer, M. and Champy, J., Reengineering the Corporation, 1993. Ir. Bas Voeten graduated as a Logistics Engineer at the Eindhoven University of Technology in the Technology Management Department. Currently he is project leader at ASM Lithography in Veldhoven, The Netherlands. Ir. Leon H. Kroep graduated in Industrial Engineering and Management Science. He is a lecturer at the Eindhoven University of Technology, Faculty of Technology Management. He has 25 years of experience in the operations and project-management eld, both as a lecturer and consultant. He has contributed to several publications about project-management and project-management systems. 188