 |
May, 1998 ADRonline monthly will be an evolving and eclectic resource that is intended to link technology and technique. We hope that it will occupy your attention for about a half hour each month. We are happy to consider recommendations to include in ADRonline monthly but please remember that our ADR Links page is a more comprehensive resource of Web-based materials. This month, take a look at: Closing the gap between ADR and technology
Online disputes and online dispute resolution
Reading of the Month
Op-Ed/Commentary/Essay/Project of the Month
|
| As an engineer, I didn’t expect to be the inventor of a product like
One Accord. Traditionally, engineers are involved with applying the exact sciences, and
avoid unpredictable processes like trying to mediate disputes and getting people to agree
with each other. So how did I get into this business? My early career focussed on water resource development. Circumstances found me in Nepal where I became involved in a very exciting and unconventional integrated rural development project. It promised to bring significant benefits from hydropower and irrigation to the poorest of the poor. Unfortunately, conflict among certain parties about how best to develop the project, in my view, seriously compromised its potential benefits. That left me quite disappointed but convinced there had to be a better way. Later, while brainstorming ideas for doctoral research at Cornell University, my major advisor, Professor D. Pete Loucks described the gap that often exists between engineers and decision-makers. He challenged me to find a way to fill that gap – a challenge I accepted. In my literature research, it was not surprising to learn that conflicts around water are among the biggest and most complex facing humanity. Taking advantage of my Nepal experience, my research was built around a hypothetical water conflict – a useful context for some ideas I developed for two-party conflict resolution. My advisory committee encouraged me to extrapolate the thinking in order to accommodate negotiations among many parties. My goal was to solve the problem with algorithms using optimization technology. One advantage of using optimization is that programmers can easily build in the constraints required to describe feasible solutions. It also gives you the option of using linear programming, which has superior ability over non-linear methods in solving problems with complex constraints quickly and reliably. The first requirement in applying optimization technology is to set objectives. So the question became: "If our goal is to help multiple decision-makers with conflicting objectives come to an optimal agreement, how will we know when we’ve reached it?" This requires a method for weighing alternatives in order to decide which one is better. We decided that, in a multi-party problem, there were two main objectives: -- maximize total benefits and do it in the fairest manner. So how do you build an algorithm that calculates the fairest way to solve a conflict? Impossible, you say? To be sure, no one short of God can make a decision like that. However, if the parties themselves can agree on how benefits should be divided, then it does become possible to craft an algorithm capable of apportioning maximum benefits for all concerned. Another challenge in this research was finding ways to elicit and model party satisfaction. The natural way of solving this problem is to take advantage of the skills already in possession of the negotiator. My reasoning was that a qualified negotiator must be able to compare one alternative with another. While it may be difficult to quantify how much better one alternative is than another, it is relatively easy to identify which one is simply better. When a negotiator does this, s/he is also able to identify equivalent alternatives, i.e. alternatives that would yield equivalent satisfaction. This is what led to developing algorithms that could analyze a party’s preferences and build satisfaction functions based on a set of equivalent alternatives. Perhaps the most difficult aspect of solving multi-party negotiation problems has been dealing with interdependencies, i.e., situations where preferences about decisions that could be made on one issue depend on decisions that will be made on another issue. I discovered that some interdependencies could be handled if a party was willing to change assumptions and identify relevant preferences by following an iterative process. Later, I also discovered how to model IF/THEN constraints with linear programming. Current research indicates that some types of interdependent multi-party problems may be better solved with non-linear optimization. My 1993 dissertation described the algorithms required to solve multi-party water resource development conflicts. However, it was apparent that if these algorithms could solve water conflicts, then they could also be applied to virtually any kind of negotiation problem. And that’s when the entrepreneurial juices began to flow about the commercial possibilities of such a product. The company I founded conducted several more years of research and development and was fortunate to recruit a brilliant young man named Ian Upright who had been working on better ways to share data between networked computers securely and efficiently. Combining our skills, we have been able to produce a software product capable of supporting secure negotiations among parties located anywhere in the world through computer terminals connected to the Internet. One Accord is both a product and a process now ready to launch in the marketplace. It attempts to bring multiple parties with conflicting objectives to an equitable (fair for all parties) and efficient (no value left on the table) solution. We believe that One Accord can contribute significant value to principals and intervenors in complex negotiations by helping them to reach better agreements and reduce the time, cost and stress involved in getting there. For more information on One Accord, visit http://www.oneaccordinc.com. |
