PM software forum
addressing software needs for resource management
The Purpose of “Concerns of Project Managers” is to provide expert knowledge and opinions on topics of general and continuing interest to PMJ readers.
The responsibility of the Feature Column Editors is to assemble this knowledge and opinion, verify it and present it in an interesting and useful manner. They are dependent upon your input, both as to topics you need information about and providing information you have. Please communicate your needs and experience. We will try to be responsive. If you have relevant information, please share it with us.
To communicate with any of the feature column editors, just write:
Project Management Journal, ATTN: [Name of Feature Column], School of Business, Western Carolina University, Cullowhee, North Carolina 28723.
The opinions expressed in these columns are those of the respective author. They are, in no way, to be construed as official positions of PMI on an issue or endorsements, either positive or negative, of any product mentioned herein.
Feature Editor: Harvey Levine
A Historical Perspective
In the thirty year history of project management software we have seen several cycles, from diversification to sameness, in the products and platforms. At the start of each cycle, we are presented with different approaches, such as we saw in the late 1950’s, with the introduction of PERT and CPM. Later in the cycle, the differences slowly dissolve into barely distinguishable products; a blend of the supposedly best of each of the original varieties. We saw this happen again as the microcomputer became a new and popular platform for project management software, only to see the differences between mainframe and microcomputer based products become more clouded with every product introduction and upgrade.
Followers of the history of project management software will remember that the PERT and CPM protocols were each developed to solve a specific set of problems. They were developed as tools to assist in the planning and management of different types of projects in different environments. Likewise, the development of project management software for the microcomputer platform, at first brought two levels of functionality to PC users. There were products that were rewrites of mainframe programs, aimed at the sophisticated user. And, for the new user, products were developed that traded off some of the functionality and flexibility for the less demanding user-machine interface, that was offered by the microcomputer environment.
The application of recognized project management techniques, coupled with the use of microcomputer-based project management software, is being embraced by an ever increasing clientele. Once it was (nearly) the exclusive tool of construction, defense, and aerospace contractors, and the utility and process industry. Now, project management, with all of its protocols and programs, is being adopted by manufacturing, pharmaceutical, data processing, banking, legal, and numerous other disciplines.
The latest generation of project management software has developed into an exceptional array of products, providing both “power” (functionality and flexibility) and “user-friendliness” (ease-of-learning & ease-of-use). Yet, some of the new members of the project management family may find that the problems that this generation of software attempts to solve are mostly the same, and that such software does not address itself to the needs of some of these new users.
We all know that three of the major areas that are managed in the project environment are schedule, resources and cost. At least, it is for these three components of project management that we look to project management software for automated assistance. But, let’s look at how this is addressed by most of the current products. Traditional project management software is based on critical path schedules. The resource and cost components are driven by schedule component. This integrated, schedule-driven approach is characteristic of classic project-management, and will serve most project management situations quite well. However, there are numerous applications that require a stronger emphasis on the resource or cost control aspects of the projects, which are integrated with, but not driven by, the schedule component. We are beginning to see a growth in the availability of project management software that addresses these needs, but, in general, users and vendors admit that we are far from fulfilling the needs and potential in this area.
In the resource area, there are programs that deal more effectively with discrete or unique resources, as well as the more traditional pooled resources to be applied to tasks for less that the resources’ total availability, and for less than the tasks’ total duration. Some of these programs have provisions to allow the resource application quantity and rate to actually determine the task duration (ViewPoint, Harvard Total Project Manager II, Project Workbench, TimeLine 3.0). Approached differently, but with a similar effect, AMS Time Machine and SuperProject Expert will recalculate a task duration if the resource application quantity and rate do not support the duration that was input. An additional feature, supporting real-world conditions, is the ability to interrupt a task to divert a scarce resource to more critical work (MicroPlanner, SuperProject Expert). These kinds of conditions are the norm in the more service-oriented industries such as the software development and data processing areas. Such industries are becoming a major source of new project management software users. There have been some recent product releases that specifically emphasize support for MIS applications. These include MicroMan II (PC) and Multitrak/Multitrak PC (IBM mainframe & PC-AT). Digital is now marketing their VAX Software Project Manager, which been developed for their own in-house use. As with many commercial products that were derived from in-house programs, it may take some time for this product to reach the full maturity needed for a wider base of users.
Even traditional users will find it difficult to accept the results of the automatic resource leveling functions that are available with the majority of project management software packages currently on the market. All resource leveling programs are based on a predetermined or user defined set of ordering criteria that tell the program which task to schedule when there is a conflict for scarce resources. There are two problems with this approach that most often results in a less than optimum solution. First, with the exception of the two “activity-splitting” programs, noted above, the traditional resource leveling programs will not permit a task to be interrupted once it is in progress. The result is a leveling computation that may leave unused periods of time, because these available windows are not long enough to allow a performable task to start and finish. Second, the use of a single prescribed ordering of tasks for leveling is more stringent than real world conditions. Today’s automated resource leveling systems, even those that allow the user to specify the ordering criteria, will almost always produce a leveled schedule that has a longer duration and less smooth resource usage than the ultimate solution that might be produced if the project manager could perform all of the computations by hand.
Some Expert Opinions
In a previous issue, we aired these problems, and promised a more detailed discussion and additional views. The PM Software Forum decided to address these issues to some of the consultants and software developers that are active in the project management arena. Participating in these interviews were: Richard E. Murphy, Executive Vice President of K&H Professional Management Systems, Wayne, PA; Andrew Layman, Vice President - Product Development, Breakthrough Software Division of Symantec Corporation, Novato, CA; Jim Dunnigan, Product Manager, Microsoft Corporation, Redmond, WA; and Ira Bitz, President of Ira Bitz & Associates, Ltd., Chevy Chase, MD. Mr. Murphy is responsible for K&H project management products such as Prestige, Cue and Premis. Mr. Layman is the author of TimeLine. Mr. Dunnigan is Product Manager for Microsoft Project. Mr. Bitz has been a recognized project management consultant and trainer for over 17 years.
* * *
PM Software Forum: In view of the preceding discussion, why is resource scheduling and leveling a problem and how can it be improved? Also, do we need a new project management software approach to resource scheduling, for a wider base of applications?
Layman: Computation of a critical path schedule is relatively simple. You have tasks and dependencies; and one right answer. The computer can do this quickly and easily. Resource management is more complicated. People have conflicting priorities, differing talents, distinct schedules, and there is no hope of computing an optimal schedule quickly enough to be of much use. So computers solve the easier problem first, that is; the scheduling constraints. In order to achieve a reasonable turnaround of resource scheduling, especially on a microcomputer, requires an efficient pairing of computer calculations and human decisions.
It goes like this. The human tries an experiment; the computer shows the result; the human tries a different scenario; and so on. The fast response and interactive use of the personal computer allow the project management program to produce results far in excess of the explicit rules built into it.
As more powerful computers are available to managers, more of their capacity will be devoted to calculations that go beyond critical path. In the future, the user will be able to enter more rules and policies about the allocation of resources, and will make distinctions between pools of people and a partial use of a person’s time, support distinct work schedules for different people, and will track results more finely.
But this does not alter the essential advantages of the personal computer. Even with the pending advances in hardware and software, the whole idea remains to make the human and the computer a team. This means that a substantial amount of the program capacity must be devoted to helping the user quickly understand the project details and the effects of changes. It must support a process like: “If I change this, what happens? Do I like the result? If I don’t, back it out and let’s try something else.”
Bitz: Yes, I agree that there are problems in the resource and cost scheduling area, but I am not sure that there is a sufficient conceptual understanding of the problem to allow for the development of systems which are not primarily schedule based. I completely concur with your comments on the resource scheduling problems, but would add one more element. Resource hierarchies are needed, so that the resources can be treated as a pool of similarly skilled individuals or as unique individuals, within the same project, or certainly, within the same database. But there are some resource applications which will require more use of artificial intelligence than is practical today. Al may allow us to alter the estimate based upon which individual in the resource pool is assigned to it. Al may also recognize that individuals have many skills and that their proficiency varies depending upon the type of work to which they are assigned. This is, I’m afraid, quite a bit off in the future.
Murphy: This is an area that, I believe, is going to change the project management software industry dramatically, more than any period in the past. Basically, I think that there will be totally new types of software being applied to the problem of project management. The PC revolution has spawned a lot of interest in project management from managers who know they should have better project control and think they can do it since software is available on a PC. Only, existing software forces them to approach project planning and tracking in a manner that is familiar. The new types of software will not require such an adjustment. I believe that we will begin to see the emergence of software, perhaps not based on Critical Path Method, with Artificial Intelligence features. One prominent software vendor, not presently in the project management field, recently approached K&H about working jointly on such an approach. We are pursuing this, based not only on our “gut feeling,” but also on market research, done by the vendor, showing project management to be a significant growth field for the right product approach.
Dunnigan: More refined resource management capabilities have been a key component of new project management software. Users have demanded from the software the ability to accurately define exactly how the resources are assigned to the activity. Developers have responded with such features as: resource calendars, partial allocation of resources, resource assignment lags, resource driven scheduling, and other such “fine-tuning” features. Resource leveling still needs much more improvement and until an “expert” is designed which can really allow the manager to optimize his resource allocations, it will still be used only as a good “what-if” tool.
Multi Project Scheduling
PM Software Forum: Despite the general concept of the microcomputer as a “personal” computer, most users and potential users of microcomputer-based project management software are specifying the ability to aggregate resources across multiple projects, without losing the integrity of the individual project. How prevalent do you see this demand and what is being done about it? Many vendors (at least 50% of them) claim to have a multi-project capability, but do they really meet user needs in this area?
Dunnigan: The issue of project management software supporting a multi-project environment, sharing a common resource pool among these projects, is very important. Software developers are addressing this need today and this capability will soon be standard in all systems. PC project management software was originally designed with a single project environment in mind. This has changed. Developers are now designing future programs with a multi-project environment in mind. Especially important is the ability to share and monitor resource usage across projects.
Bitz: I disagree that half of the micro-computer based project management software can respond to the need to integrate resource information across multiple projects. Many of these products can be “jury rigged” to aggregate data, but were designed as single project systems. They have to be told that multiple projects are part of one big project to produce such aggregations. Only a few products do this without jury rigging, and some of them have such low capacities that they are virtually useless for all but the smallest organizations performing projects.
Layman: I do not believe that any of the existing project management programs (TimeLine included) do nearly as good a job at managing resources as they could. When faced with a multi-project environment, each of the programs represents compromises and opportunities for improvement.
Coordination of multiple projects requires technical sophistication, and also a recognition of sensitive human and political factors. Among the problems to be solved are:
- Who owns the data, the line manager or the central data administrator?
- How do we prevent innocent what-if analyses from polluting other related projects?
- How can software help to arbitrate conflicting demands on resources? On Milestones?
- How can software ensure that all data reflects the same reporting cycle?
* * *
It is most comforting to see the response of these perceptive project management experts. It should give us the feeling that the problems of resource management are being recognized, and that considerable thought is being given to pragmatic solutions. The PM Software Forum thanks these project management practitioners for contributing their insights to this complex and vital subject, and invites its readers to add to the dialogue.
Georgia’s on My Mind for PMI’89
PLAN YOUR VACATION AROUND PMI’89
Did you know .....that the Nation’s first gold rush was ion the State of Georgia? In Lumpkin County, in 1828? And that gold from that site was used to plate the State Capitol Building dome in Atlanta?! Come to PMI’89 and see the gold-plated statehouse roof — one of only ten in the U. S. like this.
THE PM NETWORK November 1988