Products & Services
Flow-Works Manufacturing Software
Flow-Works software supports several critical functions for effective management of continuous flow (lean), mixed-model manufacturing. Mid-volume products with customer options that cause assembly crew balance variations are especially good applications.
Flow-Works functions under an ERP or MRP umbrella system with MRP performing the critical role of managing the materials supply to manufacturing operations. Then Flow-Works is used to plan and manage the fast-paced transactions within flow manufacturing lines where real-time tools such as Kanban visual signals are needed to keep pace. Flow-Works linkages to the ERP system are limited to an essential few (e.g. BOM and order file downloads) thus avoiding redundancy and simplifying implementation and maintenance.
Flow-Works supports three key decision-making phases of mixed-model flow manufacturing:
What Makes It Unique?
Flow-Works uses the power of next generation autonomous agent software to manage the millions of naturally complex product and process combinations that occur in mixed-model flow manufacturing. Its beyond object-oriented architecture provides real-time answers to management problems that conventional systems cannot supply.
Simply-stated, Flow-Works is a data manipulation tool set that uses the power of autonomous-agent parallel processing for accurate simulation and management of the millions of product and process combinations in mass-customization manufacturing operations. It is complexity theory used to simplify every-day manufacturing problems.Next Generation Software for Complex Systems ("Chaos Theory")
Researchers who developed complexity theory (sometimes called chaos theory) learned that:
Putting those two concepts together led to the autonomous agent computer systems concept, a leading-edge development that overcomes the limitations of conventional software to manage complex systems.Central Systems vs. Taxi Driver "Agents"
To illustrate the agent systems concept, we can use a real-world example. Consider the taxi cab system in Chicago. The primary objectives of the taxi system are to maximize passenger utilization of available cab time thereby maximizing revenues while minimizing overall taxi system costs of operations.
Now suppose we want to install a planning system to improve overall taxi system performance. We could take one of two approaches. A traditional computer system would look at where taxicabs picked up passengers last year, perhaps take into account hotel meetings and other events, and then tell all of the cabs where to go. An alternative would be a central dispatch function with 100s of operators taking calls and then sending the cabs to a specific location.
You can guess the probability of success of both scenarios.
To be effective, the computer system would have to accurately simulate and plan for the aggregate of millions of interacting behaviors of taxis, passengers, public events, weather and other factors. But a few simultaneous equations in conventional software systems cannot accurately simulate those collective behaviors. At best, the results can be only rough approximations of the taxi system.
People have tried central systems to manage taxi operations with marginal results. In the end, taxi systems work best when governed by a few simple rules for the taxi drivers (1. Drive passenger to destination as quickly as possible. 2. If idle, drive to nearest taxi queue. 3. If the wait at the current queue is too long, drive to the next one. Etc.). Some drivers will elect to work through dispatch, some will just focus on airport runs and others will coast around town. The end result is a system of "autonomous agent" drivers that largely works.
Ants and Other Complex Systems
Many natural systems operate under "autonomous agent" rule sets that are aligned with the goals of the overall system. Ant colonies, birds, fish, human families, and community groups operate under rule sets that guide the behaviors of the individual agents for the common goals. Like the taxi system, modeling of most natural, complex systems is beyond the capability of conventional software.Conventional ERP and MRP Systems
ERP and MRP use elaborate mazes of simultaneous equations running in fourth-generation and object-oriented languages. In spite of their complexity they are able to produce only rough approximations of the complex operations they attempt to simulate for planning purposes. Inherent limitations prevent them from accurately simulating the millions of product and process combinations that are necessary for seamless operation of continuous-flow, mixed-model production operations.Future Management Systems for Complex Processes
"Autonomous Agent" software takes advantage of the massive power of computer hardware now available to simulate complex manufacturing processes, at relatively low costs. It uses a structure of autonomous agents, operating in-parallel to real-time simulate the inter-related behaviors of the operators, machines, conveyors, materials, orders and other significant "agents" within a mixed-model manufacturing process.
"Autonomous agent" software allows accurate simulation of real-world complex systems. With an agent-based system it is possible to test agent rule set changes quickly, and input changes on-the-fly like, for example, adding a rule for all cabs in a certain radius and time to move during idle times in the general direction of a particular public event site.Evolution of Computer Systems Beyond Year 2000
Conventional software suppliers have been slow to adopt autonomous agent technologies, probably because of the huge investment in their current systems. Looking ahead, and considering the pressures of global competition, autonomous agent software offers significant improvement opportunities for effective management of complex systems.
© 2009, The Cumberland Group, Inc. All Rights Reserved
© 2009, OmniCom Solutions Group