As more manufacturers in all industrial sectors embrace the techniques of the Toyota Production System, terms like Kanban, Poka Yoke and Jidoka are becoming part of the fabric of more drug manufacturing facilities. Jidoka, also called autonomation, gives equipment and operators the ability to see when an abnormal condition has occurred and to stop work immediately, building quality into each step of the process.
A key part of Jidoka is error proofing, or systems that either prevent operator error or allow prompt response. It is a process where people are asked to identify every instance where the situation doesn’t match the expectation. More equipment is being designed to eliminate operator error and improve root cause analysis, while radio frequency identification (RFID) is playing a part in more systems throughout the plant.
Source: The Five Steps to Building Jidoka Equipment, J. Miller, Panta Rei, www.gembapantarei.com, May 4, 2006.
Editor-in-Chief Agnes Shanley interviewed three experts in automation for life sciences — Bob Lenich of Emerson Process Management, Bob Honor of Rockwell Automation, and Bengt Stom of ABB — to elicit their perspectives on the current state of, and future prospects for, automation and process control solutions in pharmaceutical and biopharmaceutical operations. Their respective Q&As follow.
A 30,000-foot-view from Emerson’s Bob Lenich
PM — Looking broadly at the automation and process control solutions now available to pharma and biopharma manufacturers, would you say that today's life sciences solutions fully reflect this principle in their designs, or that they are evolving to that point? If they haven't reached that point yet, what is still needed?
BL — The principle of Jidoka is seen today in the many responses life sciences producers are making to the FDA PAT initiative. The evolution includes expansion with more reliance on in-line analyzers to provide accurate real-time feedback, applied modeling and simulation and optimization techniques to permit decision making and more “closed-loop” processes.
PM — Are end users in pharma/biopharma demanding these features in their solutions? What differences do you see in this regard between customers in pharma vs. those in automotive or discrete manufacturing?
BL — The leaders of the pharma industry are demanding and using new responses to the PAT initiative to improve their operations. There are some leaders in this field including in our experience — Pfizer, AstraZeneca and Wyeth, among others. These advances are currently being seen most commonly in the fill/finish portions of the pharmaceutical manufacturers with the goal of increasing product throughput and decreasing the cost of quality by enabling parametric release of products. In our current view and experience, biological manufacturers are not as invested in this technology primarily due to the more difficult nature of their measurements and corresponding lack of accurate instruments to rely upon.
We don’t deal with automotive industry in the main so direct comparisons are more difficult to make. It is clear that the ability of piece-parts manufacturers to measure their works-in-progress against their specifications (size, weight, roundness, etc.) is, in the main, easier than determining the exact chemical nature of a complicated pharmaceutical intermediate. It is logical that the Jidoka practices and principles that apply more broadly there would be more entrenched at this time. In a similar fashion, once an active pharmacological ingredient or biologic product has been produced and now need to be packaged. The processes are very much aligned and the Jidoka principles apply very well.
PM — Which products and platforms has Emerson introduced within the past year that best embody the principle of Jidoka?
BL — There are three main areas in which Emerson Process Management is responding to the life sciences demands for Jidoka practices.
- Abnormal Situation Prevention: This permits pharmaceutical manufacturers to avoid disruptions in their process due to predictable failure analysis of capital equipment. While not exactly tied to in-line measurement, a failure of a key piece of hardware in the middle of a batch can clearly disrupt production and waste valuable time and raw materials. Emerson has techniques that help end users minimize disruptions due to predictable equipment failure.
- Recipe Authoring and Electronic Work Flow Management: A key part of life science management is the ability to control precisely the process necessary to make product and to record accurately for reporting purposes what was done and how. This tool set provides life sciences manufacturers the tools needed to produce product repeatably and prove that they have done so.
- Bioprocess Modeling: Emerson Process Management has developed and is in the process of testing and releasing tools that permit the modeling of complicated bioprocesses. This will provide the biopharmaceutical industry with a valuable new tool that will permit more accurate and repeatable cell manufacture and active ingredient manufacture. It is another tool that moves beyond the current wait-and-test approach that is so antithetical to the Jidoka philosophy.
Perspectives from Rockwell’s Bob Honor
PM — Looking broadly at the automation and process control solutions now available to pharma and biopharma manufacturers, would you say that today's solutions fully reflect the principle of Jidoka in their designs, or that they are evolving to that point? If they haven't reached that point yet, what is still needed?
BH — All automation suppliers, but especially Rockwell, provide various means of displaying real time information and having operators make decisions on whether to stop the process or better yet, providing real-time means for the automation systems to make this determination. For pharma, however, this gets into the PAT and risk-based manufacturing realm which requires full scientific understanding of the manufacturing process. For those manufacturers embarking down the path of risk-based manufacturing and heading toward parametric release, all automation and process control suppliers have a wealth of tools to help automate that process.
PM — Are end users in pharma/biopharma demanding these features in their solutions? Are there any differences you've seen compared with regard to Jidoka in your pharma customer base, compared with your automotive/discrete manufacturing customers?
BH — End users are not demanding this capability. They are exploring the analytical and modeling tools available to help them fully understand and predict their manufacturing processes so that they can start employing these kinds of advanced process control techniques.
Many other industries have examples of fully automated production, up to and including “lights out” manufacturing, for these kinds of quality decisions to be built right into the process. The plastics industry provides great examples for in-line automated quality control. The semiconductor industry has been operating under just these principles for over 20 years due to the much less complex molecular models required in semiconductor manufacturing processes and the lack of regulatory oversight in manufacturing semiconductors. As such, there are semiconductor facilities with process capability exceeding eight sigma.
PM — Which life sciences products and platforms has Rockwell introduced within the past year that best embody the principle of Jidoka?
BH — Rockwell has been building integrated control platforms for a generation and have them operating today throughout the automotive industry where Jidoka has been most popularized. In the last year within our Integrated Architecture platform, we added a powerful instruction called the add-on instruction which allows manufacturers to embed within the controller any algorithms, models or data developed to support specialty functions like multivariate analysis or predictive modeling or statistical process control.
Interview with ABB’s Bengt Stom
PM — Pharma's adoption of robotics appears to be limited so far. In our most recent survey, we found that only 25% of respondents were using robotics (those who have adopted it appear to be using it mainly for packaging and sample preparation) and most respondents weren't even considering using robotics in the future. What are the challenges involved and why is this mindset continuing? What will be needed to change it?
BS — That pharmaceutical companies have been reticent in addressing the use of robotics stems largely from the recent history of producing very large volumes of similar product 24/7 with no need for flexibility. Such production often lends itself to fixed automation whereby recipes never change. However, pharma companies will likely never see blockbuster drug products in the future, flexibility in both volumes and packaging types, will increasingly pressurize manufacturers to introduce ever greater flexibility. This is where robotics technology always beats fixed automation systems. Of course, there is also the perfectly normal human prejudice of ignoring the unusual or unfamiliar because it challenges conventional thinking and experience.
PM — How is ABB incorporating the idea of Jidoka (as articulated in the Toyota Production System) [http://www.strategosinc.com/jidoka.htm] in its life science product range, specifically in robotics?
BS — There is widespread confusion over how to define Jidoka. Some interpret this as simply the ability to automatically stop a system if things go wrong. ABB prefers the more holistic concept that the automation system is intelligent enough to assess what is going on and act accordingly — rather as a human might. To this end, ABB has addressed the issue through a combination of software and hardware. By incorporating vision systems, weighing transducers and proximity sensors, the robots can be instructed to act in accordance with the performance of the overall system. In other words, if a line slows, or product flow falters, the robots simply adjust their actions accordingly. The key to the integration of this proven hardware lies in the ability of the software to integrate the robot controls with the senor data provided. This software is also made simple to use in order to open up the possibility for increased machine intelligence to everybody.
PM — How far is the industry from embracing the concept of artificial intelligence in its facilities?
BS — Again, the question begged is, “What constitutes artificial intelligence?” If this simply infers the ability of a machine to make rudimentary “human-like” assumptions and decisions, then many automation systems — including those installed by ABB Robotics — already have artificial intelligence. However, if what is meant is the ability of machines to learn, then I think this is some way off yet. The solution to true artificial intelligence lies with computer scientists, ancillary sensor designers and software developers.
PM — Please comment on any products and platforms introduced within the past year that best embody the principle of Jidoka in their designs.
BS — The robots designed and manufactured by ABB Robotics are simply machines — very sophisticated, super-fast and accurate machines, but machines nonetheless. The key to meeting Jidoka principles lies in systems, and that means software and sensors. ABB is constantly developing new software and controllers to exploit both greater intelligence in the silicon hardware and increasing performance of the machines themselves.
Below is a sampling of products and services that can be applied in a wide variety of operations and are specifically designed to reduce or eliminate operator error at your facility.
Vision ID Tracks Process
The PickMaster 5.0 software for programming high-speed FlexPicker Delta robots incorporates powerful vision identification and inspection tools combined with a conveyor tracking process. PickMaster 5.0 supports installations with one or several robots working in relation to combinations of multiple stations with lines in and out. These stations deliver and feed diverse product dimensions in a large number of different picking formats.
The software serves the entire chain, from object definition to the optimal selection of palletizing patterns, building up entire stacks of pallets and calculating the best and most efficient robot movements.
ABB Robotics
RFID Validates Connections
RFID Connections are RFID-enabled couplings with added electronics to measure and identify critical parameters. The IdentiQuik series of couplings utilizes RFID technology to automatically identify fluid characteristics and capture data from point-of-origin to point-of-use.
Colder Products
Consultants Help Error-Proof Automation
Descartes Technologies specializes in error proofing and quality control systems for manufacturing.
Regardless of what automation system your plant uses, the company says it can provide solutions that integrate seamlessly with existing equipment and ensure the process is always in spec. Solutions adapt to changes in production orders in real time. Using Descartes’ EP1 all-in-one interface, complete stand-alone systems can be deployed for as little as $5900 including operator interface, PC-based HMI, bar code reader, assembly and installation.
Descartes Technologies
Diagnostics Flowmeter Features SPC
Rosemount’s 3051S advanced diagnostics flowmeter alerts operators to pending abnormal situations before they can occur. New diagnostics allow users to detect, troubleshoot and prevent abnormal situations.
The devices incorporate statistical process monitoring technology that provides warnings of such problems as plugged impulse lines, changes in fluid compositions and other events. The devices allow for variable logging with time-stamp capability and advanced process alerts.
Emerson Process Management
Eliminate Packaging Insert and Printing Errors
iCount is an automated workstation designed for non-contact counting of folded, stacked products such as outserts, inserts, leaflets and cartons. It is capable of counting trays in 1/30th of a second, its manufacturer claims, saving labor costs and improving accuracy.
Docu-Proof C2, meanwhile, eliminates insert printing errors by using text comparison technology. The system can compare text in virtually any language, and can detect change in fonts, font sizes, deletions, inserts, spelling and can even detect errors in Braille printing. It includes an extensive reporting feature and is FDA- and EMEA-compliant.
Global Vision, Inc.
Troubleshoot Your Digital Dosing
The TrueDos flow monitor provides operators with information on what’s happening in the dosing head. The device was designed to establish that the suction/pressure response in the dosing head conforms to a fixed rule, regardless of the quantity being conveyed. The device allows dosing malfunctions, including air bubbles in the dosing head, cavitation and pressure drop due to a line break, to be detected and reported immediately, even at low flows.
Grundfos Pumps Corp.
www.grundfos.comEliminate Filter Data Errors
Millipore’s SMART Technology combines a built-in RFID tag and data reader to provide accurate information about any individual filter cartridge.
Each tag stores such information as filter catalogue number, product lot number, serial number and unit-specific diffusion value. The system is designed to allow end users to monitor filtration systems in real time, and to access and store process information via wireless systems. SMART Technology’s aim is to eliminate tedious manual data reading, recording and transfer. Its manufacturer plans to add the following features:
- Incorporation of wireless RFID temperature sensor enabled cartridges to improve autoclave and SIP filter validation
- Improvement of inventory management through recording of each part number during inspection in the receiving area
- Reduction of errors entailed in exchanging data with DCS
- Confirmation that the proper filter is being installed by checking the RFID code with an existing MES.
Millipore Corp.
Traceability and Error-Proofing Expertise
ToolWorx Information Products’ TWLot is a serialized lot-tracking system for raw materials to finished goods. It includes bar code and RFID scanning of components linked to containers or equipment or pallets, collection of test data from programmable logic controllers (PLCs), error-proofing to ensure that the right components are brought to each work-cell report module for part and lot geneology.
