Transforming how we model material handling and logistic systems

By Benoit Montreuil, President 2008-2009
College Industry Council on Material Handling Education (CICMHE)

Modeling is key to everyone in the material handling and facility logistics design community. Equipment manufacturers develop functional and technical models of their equipment for research and development, sales and estimation, production and supply, reliability and maintenance, and so on.

Every software firm providing execution and control systems has to develop models for its software development and integration with physical and management systems. Every system integrator and consultant develops models of the system to be designed and integrated to insure that, beyond its components, the system as a whole will be performing as expected by the client. Every heavy user needs models to plan the evolution of its systems and facilities so they keep on being business enablers.

The New Modeling World
Yet the world is moving and we have to adapt. Modeling as we have done it up to now has to be reinvented. This reinvention has to be on at least four interlaced counts. First, we must elevate our modeling perspective. Second, we must stop reinventing the wheel. Third, we must learn to share and thrive from openness. Fourth, we must embrace persistent modeling.

Elevated Perspective
Regarding the first count, Professor Detlef Zulkhe brought an interesting perspective in his keynote address at the 2008 Congress of the International Federation for Automatic Control held in Seoul in early July. He pointed out that in the 80s, the electrical engineers went away from modeling in terms of transistors, gates and circuits to thinking in terms of functions and systems.

In the 90s, software engineers went away from writing lines of code to model in terms of objects, systems and agents. This decade, mechanical engineers go away from delivering drawings to modeling mechatronic objects and systems. The same transition is occurring in useware engineers, going away from placing buttons, switches and displays, toward modeling interactions between humans, machines and computers. Dr. Zulkhe advocated the same type of elevation for control engineers. We need such modeling elevation in the material handling and facility logistics community.

Stop Reinventing the Wheel
On the second count, we mostly do modeling like artisans, starting from basic raw materials and shaping situation specific models. Even within a single organization, whenever a new design challenge is posed, engineers work at studying the problem and building the model. Most often, previously generated models by other members of the organization are not even looked at and considered as modeling bases.

In the paper I presented at the 2006 International Materials Handling Research Colloquium, I advocated that we should get away from the paradigm that our case is different from everyone else’s case and recognize that most production and distribution centers in the world have much more in common than they have differences. I showed that using a limited set of protomodels such as producers, assemblers, processors, distributors, fulfillers and transporters, one can describe most operational centers in the world.

For each protomodel, there are a number of design specifications and expectations which are common, simplifying, speeding up and guiding the modeling work. Such proto models could be developed for most materials handling system types.

In fact, as Professor Pierre Lefrançois and I proposed as early as 1993 in the Journal of Intelligent Manufacturing, each manufacturing and handling equipment sold by a solutions provider should be delivered with its own virtual model certified to document and represent it as best as possible, so that all concerned have a ready tool for helping select, implement, maintain and control it.

At the IFAC 2008 conference, Professor Zulkhe similarly advocated that all physical devices should come with a device model encompassing a service model, a product model, a communication model, a CAD model, and so on.

Embrace Openness
The third count is about stopping to consider models as strictly confidential, to be hidden in specific organizations, but rather to expect them to be rendered as open as much as possible so that we can all build on each others’ shoulders, sharing our expertise and accomplishments. Software engineering is undergoing a worldwide sweeping revolution with the Open Source movement. We should do the same in our community and create an Open Model movement.

Yes it is tough to think that the model we have spent six person-months to create should be made available to the world, but think that if somebody else had done the same earlier with their model you could have done yours in a few weeks, learning from theirs and exploiting worthwhile components. Yes, confidentiality of data is important, but it is possible to make sure that key data is transformed so as to avoid compromising clients’ competitiveness. The power of sharing is worth reconsidering our ways of doing.

Persistent Modeling
Finally, on the fourth count, we must get away from throw-away-after-usage, designed-to-be-discarded modeling, to embrace persistent modeling which is exploited along the entire life cycle of the modeled system. For example, it is common for a solutions provider or an integrator to develop a simulation model of a complex conveyor system. Yet this model is most often not transferred to the user. So the simulation model gets stored in the vaults of the provider or integrator, sitting there unused, until perhaps there is a new request from the client. There is so much waste involved in this behavior.

The model should be designed from the ground up knowing that the user will exploit it for projective throughput and reliability purposes, for analyzing and improving controls or physical components. The model should make it easy to continuously feed it with up to date information, and for it to provide engineers, technicians and managers with an easy to use and powerful tool for multiple purposes during the system’s life.

On each count, members of our community have made significant advances. Yet there is much to be done. The 2007 Material Handling and Logistic Summit has identified Comprehensive modeling and unified digital representation for Material Handling and Logistic systems and solutions design as one of the key initiative themes to be undertaken by the community. CICMHE members have taken leadership along this theme.

Indeed professors Leon McGinnis and Russ Meller have recently been awarded NSF funds to investigate the feasibility of an Engineering Virtual Organization for Discrete Events Logistics Systems aiming to network all researchers involved in our community. This initiative is generating significant enthusiasm in the academic research community, as validated at the recent 2008 International Material Handling Research Colloquium.

The type of evolution I here advocate will require the openness, engagement and leadership of individuals and organizations all across our community. Industry has to take a major proactive role in this evolution, or it will remain academic dreaming.

For example, if every company would identify one key person, an experienced modeler or executive, to act as catalyst in its organization and networker with the rest of the community, we could create momentum with the potential to really transform in a few years the way we all do modeling.

Think about it, discuss among yourselves, and if it makes sense then please act upon it.

<< Back to news articles