A Pilot Project

1
A Pilot Project

• Cautiously Moving Forward with an Industrial
  Wireless Initiative in a Process Research and
  Development Facility

2
Presenter

• Jim Murphy - With a B.S. degree in chemical
  engineering from Purdue University, Jim learned
  process control working with Fisher Controls and
  Measurex systems at Eli Lilly and Company. After
  spending another 10 years as an application
  engineer for ABB and Foxboro, Jim returned to Eli
  Lilly and Company in the late 1990s. In his
  current role, Jim is charged with implementing
  automation solutions in a process development
  environment. Wireless is seen as a key strategy
  for the rapid deployment of portable processing
  equipment, and the flexible placement of campaign
  specific instrumentation.

3
Current Situation of Wireless atEli Lilly and
Company

• Corporate infrastructure group supports wireless
  extensions of business LAN for café-style VPN
  connectivity. An access token provides
  additional security to normal user
  authentication.
• Traditional wireless solutions are utilized in
  warehousing.
• Point-to-point wireless solutions are in place to
  communicate across geographic boundaries within
  utility systems.
• Wireless is not utilized in manufacturing within
  control systems.
• Investigations for RFID and asset management are
  ongoing. Nacer Hedroug, a Lilly associate, is a
  co-chair of the SP100 RFID SG.
• From time to time, rogue wireless access points
  are utilized that violate corporate security
  policies.

4
Current Situation within Our Plant

• Wireless access to business LAN is available in
  several of our conference rooms.
• Wireless infrastructure has been installed in our
  pilot plant to support future in-process
  information collection (deployment is the same
  model as meeting room access).
• Point-to-point wireless solutions regularly used
  connectivity between PAT, and hosting laptop
  devices.

5
How we got here???Our Business Case

• Our facility executes development and cGMP
  batches of early phase compounds.
• Development information can be as important as
  the material being produced.
• Our development scientists and engineers want
  more information, and fixed installs of adequate
  instrumentation are not a viable solution. Ad
  hoc placement of various instruments is
  desirable.
• More and more the process control group is being
  asked to provide integration and monitoring
  capabilities for new and proposed portable unit
  operations equipment. Our distributed control
  system is our focal point for integration with
  fixed processing equipment and connectivity to
  our process data historian.

6
Pilot Plant is on Three Floors150 x 80

• Graphic below depicts the existing coverage of
  installed wireless.
• 1st Floor 2nd Floor 3rd Floor

7
Investigation of Alternatives

• To support the ad hoc deployment of single
  instruments, we have considered taking advantage
  of existing infrastructure for ad hoc
  instruments. There is existing, I/O bus wiring
  to mil-spec connectors in several work cells.
  Adding some combination of new wire, compatible
  I/O cards, and software in each work area is
  possible, but is does not meet our needs for all
  areas.
• Fiber is in place today that could be used to
  provide Ethernet to almost every area, but is
  intended future DCS workstation replacements.

8
Investigation of Alternatives(continued)

• Our DCS requires workstation upgrades to be kept
  in a supportable state. The preliminary upgrades
  being conducted this year enable new I/O types to
  be incorporated into our system. The capability
  to connect via Ethernet to our PLCs and OPC
  connectivity gateways have created the most
  interest.
• Use of existing wireless access points has been
  considered, but usage requires MAC address and
  port mapping via change control. Given the
  desired portability, this is not favored.

9
Perfect for Wireless

• Anticipated instruments are intended for
  development measurements, and are not considered
  mission critical.

10
Perfect for Wireless (continued)

• Connectivity to portable equipment will provide
  gateway for monitoring and data historian
  connectivity. A local HMI will be provided, and
  no control or event monitoring will be done via
  the DCS.

11
Perfect for Wireless(continued)

• No loop closure will take place over the wireless
  link.
• As a pilot facility we are a logical venue to
  trial new devices and applications.

12
Wireless Industrial Network Std We will limit
our implementations to Classes 3 - 5
Reference ISA-SP100.14 CFP, WW.ISA.ORG
13
Our Design

• Level 1 or level 0 wireless network
• No possible way to hack into business LAN
• Flexible multi-radio network appliance(12 15
  required)
• Wireless mesh with 5 GHz backhaul
• Ethernet access (802.11b/g)
• Instrument connectivity via 900 MHz Wireless HART
• OPC and Ethernet I/O cards on DCS

14
System Overview
15
Wireless HART More Detail
16
Wireless PLC Comm... More Detail
17
Where We are Today?

• Our project has acceptance within the process
  control discipline, and among our users.
• Communication to IT infrastructure group has been
  tedious.
• Fortunately the system owner is the same for the
  future application, the installed wireless
  hardware and our proposed industrial wireless
  network.
• A few documentation gaps remain, and we are
  working with our vendors and IT representatives
  to address outstanding issues.

18
Lessons Learned

• Begin the dialog with IT governance organizations
  early. We did not anticipate concerns around
  bandwidth protection and security of the
  installed system.
• Project consolidation opportunities are not
  always obvious.

19
Acknowledgements

• David Adler, Jeffrey Owen and Daniel Heighway,
  Eli Lilly and Company
• Hesh Kagan, Invensys Process Systems
• Dr. Peter Fuhr, Apprion Inc.