In Dow Chemical's Adoption of RFID, Many Lessons

By Mel Duvall  |  Posted 2007-09-17

While many companies are still determining whether they want to make the leap into applying radio frequency identification (RFID) technology to their business practices, Dow Chemical is wrestling with the enviable challenge of having to whittle down opportunities.

A call two years ago for RFID project proposals among the company's wide ranging businesses generated more than 450 responses. Dow is already producing concrete results from many of those proposals, including using RFID technology to track hazardous goods shipments, to monitor temperature, pressure and the general condition of individual containers, and even detect termite infestations (see "An Application That's Buggy by Design").

In fact, Midland, Michigan-based Dow has developed one of the more advanced programs for not only identifying how RFID can best be applied to its wide ranging businesses, but also for capturing and sharing best practices.

Back in 2004, when Wal-Mart was detailing its far-reaching plans for using RFID throughout its supply chain, Dow Chief Information Officer Dave Kepler quizzed his team about whether the company should be looking more seriously into RFID. Dave Asiala, Dow's information technology director, recalls that at the time the general feeling was that RFID wasn't ready for prime time.

Kepler was more optimistic. In July of 2005 he decided to put out a call across Dow's far ranging operations-the company has annual sales of $50 billion in areas like chemicals, plastics and agriculture-for proposals on how RFID might be applied. His hunch, that RFID could answer a number of existing business challenges, generated close to 450 proposals.

The next step was to evaluate those proposals, based on the company's Six Sigma methodology, to find 10 projects to approve. "There were many more that had merit, but we always have more ideas than money to spend," he said.

In conjunction with the evaluation process, Dow established a center of expertise dedicated to RFID. It would be the responsibility of that center to not only identify the best projects to proceed out of the initial 450 proposals, but to also evaluate emerging RFID technologies and capture best practices. "Basically, when a business unit has a need that might be solved by RFID, we're available to them as consultants," said Craig Casto, RFID Global leader.

In order to narrow the initial 450 projects down to a more manageable number, the evaluation team applied a fairly broad vetting process. It looked first at the project's likelihood for success, based on the readiness of the RFID technology being applied, and the challenges involved, and then at the total value of the project, both tangible and intangible. By plotting the projects on a matrix, Dow was able to narrow the list down to 50 projects.

Next, Dow decided to reach out to its information technology partners and the RFID vendor community to conduct a more practical analysis of the remaining 50 projects. It essentially created a technology advisory board comprised of representatives from such companies as SAP, Cisco, Intel, Savi, and Accenture to evaluate the proposals with an eye to which projects could provide the most benefit and which were strategically aligned with corporate goals. Asiala said the advisory board was crucial in identifying projects where standards were not ready or the technology was not mature enough to move forward.

An example was the use of RFID to help locate and identify parts used at a construction site. At a construction site, parts are typically stored in a "lay-down" yard, essentially a large area where they can be retrieved as needed. It was thought RFID could be used to keep track of the parts and speed their retrieval. Asiala said while the idea has merit, and a potentially valuable payback, the technology wasn't ready yet. The project would have involved setting up a network of RFID readers or sensors over a large area. "We felt the technology needed to mature a little, but those technologies are starting to come out onto the market," he said.

After a roughly six month evaluation process, the green light was given to 10 RFID projects in the following areas:

's 10 RFID Projects">

Dow's 10 RFID Projects

  • Cylinder tracking-using RFID tags and readers to monitor the transport of gas cylinders.
  • Tanker truck tracking.
  • Rail car tracking and product monitoring.
  • Intermodel container and content tracking (containers that can be easily transported from trucks to railways to ships).
  • Warehouse and inventory management. (This lets Dow locate and re-direct containers in transit. For example, if a customer cancels an order and another customer orders the same product, it can be easily located and shipped to the new customer.)
  • Product consignment inventory management (leveraging information collected from railcar tracking).
  • Pipeline worker monitoring. (Workers have an RFID device on their belts. If the device goes horizontal, indicating a fall or injury, it triggers and alarm.)
  • Tracking inventory in the company's Agro Sciences unit
  • Hazardous material tracking
  • Maintenance parts tracking

In some respects, RFID provides Dow with an ideal solution to some of its more vexing challenges. Federal regulations require Dow to closely track and monitor shipments, particularly those involving hazardous materials. More advanced RFID tags, which contain sensors to monitor such things as temperature, pressure, and vibrations (indicating that a gas cylinder or chemical drum may have been ruptured), provides Dow with ways to automate and improve its monitoring capabilities.

Dow worked with Savi Technology of Mountain View, Calif., for example, to implement its SmartChain offering to improve hazardous goods tracking. Savi, which offers SmartChain on an application service provider (ASP) basis, uses active RFID tags to monitor rail cars. As the rail car passes various readers installed along tracks, SmartChain matches information transmitted by the RFID tag against expected routes and destinations and can also monitor conditions on the rail car such as temperature and pressure.

The system can identify whether a rail car has been opened and issues alerts when appropriate. If it identifies, for example, that a hazardous chemical rail car has been opened on a secure Dow site, no alerts would be issued. But if the car is opened at a non-protected, or non-scheduled site, an alert would be issued.

"The value that technology provides us is really critical," said Asiala. In fact, where RFID systems have been deployed, Dow has achieved a 50 % improvement in its ability to identify and resolve in-transit problems. Other notable benefits include a 90% improvement in the company's ability to offer delivery time windows, a 20% reduction in excess stock inventory, and the elimination of a historical 10 to 15% human error rate associated with the manual work processes to capture and enter data.

Insights from the RFID projects undertaken to date have been gained on many levels.

For one, Dow has used the projects to help it establish a catalog of Most Effective Technologies or METs which can be leveraged by the business units. As RFID Global leader Casto said, RFID technologies, whether tags, readers or software, are evaluated under three levels of MET.

At Level 3 are technologies that are good for a particular geography or location, such as using an application to track tools used at a single Dow plant. At Level 2 are technologies that are good for one business, but that can be utilized at multiple locations. An example is an application for labeling Dow Styrofoam insulation boards, which would be unique to Dow Building Solutions.

And a Level 1 RFID technology is a technology that has wide applicability for many business units at most locations. Dow's container tracking application, which tracks all container types and is implemented globally, is an example of a Level I MET.

Another key lesson learned, said Asiala, is that RFID isn't necessarily the right or most cost-effective answer to a problem. Sometimes a simple bar code will continue to be the best approach. In other instances, Dow has discovered the best approach involves combining RFID with multiple technologies, such as bar codes and GPS systems.

"We're willing to be innovative when the business case is there," said Asiala, "but there are going to be times when the advantage of using a passive (RFID) tag over a barcode is insignificant."