Dr. Hentz also directs the contract analytical services program at BTEC. Prior to BTEC, he spent nearly 12 years developing high-throughput screening assays, automating and optimizing laboratory equipment, and investigating new technologies with Eli Lilly & Co. and Bristol-Myers Squibb. Dr. Hentz has also served as an independent consultant in the automated liquid-handling quality control, microfluidic separations, and ADME/tox fields. He received his PhD in analytical chemistry from the University of Kentucky in 1996, and his BS in chemistry from Eastern Michigan University in 1990.
Q: Can you provide some details on what your group is involved in?
A: Here at the Biomanufacturing Training and Education Center at North Carolina State University, I run the analytical group, and we provide support to our biomanufacturing operations, teach students, and also perform contract work. BTEC is a manufacturing facility, which was set up back in 2007, to train students for the biomanufacturing industry based here in Research Triangle Park. Our primary objective is to train students, and we teach them everything from cell culturing techniques to scaling up (to 300 L) using different types of bioreactors (disposable and stainless steel). We also teach them the downstream processes that include harvest, cell lysis, clarification, and purification, right up to bulk fill. My group supports everything on the analytical side, including in-process testing (concentration, purity, and safety) and protein characterization.
Q: The key question for most lab managers is, do I really need automation?
A: I have lived through that question, and it’s really about managing the instrumentation usage. Most people use instruments, but nobody has a clue as to what the percent use is on those instruments. You first need to understand how frequently the instruments are being used, what they are being used for, and when they are being used. If processes are not automated, time is often wasted in the overnight hours or early in the morning when the instruments are sitting idle. Automation can be set up to run 24/7, but lab managers first need to understand where the inefficiencies are in their processes. If there are gaps in the process, then automation can come in to improve the efficiency, and this can be done by either partially or fully automating the process.
Q: What if you already have automation in place but it needs to be replaced or upgraded?
A: If you already have automation and need to know if it should be replaced, then you have to do the right analysis to pick up the failure modes. What you do there is comb through all your log files, using some informatics tools to help. You monitor the data over time and find the mean time to failure. When the time gap starts to shorten and you have more frequent occurrences of instrument failure, that’s a clue to look into the system. If the problem can be fixed, it’s fine, but if you have to keep fixing frequently, then it’s probably time to replace the automation.
Q: How do you decide whether to fully or partially automate your process?
A: I would say you should do what makes sense. You may not have the people or the workflow to justify 24/7 automation. You can pick an option, such as investing in a liquid handler that can meet your intermittent automation needs. The fully automated workstations [that] can be attached to different modules can be very expensive. Instead, you can automate a small portion of the workflow or use automation to replace a person doing repetitive tasks, such as pipetting, to improve reliability at a much cheaper cost. This is something I call a stripped-down automation platform. This concept has been around for a couple of years now, and it certainly seems to add value to a screening group in academia and in small companies.
Q: What are some of the disadvantages that creep in due to automation?
A: One of the disadvantages is around reliability, because people try to make automation work for everything. Sometimes these systems are just not appropriate, and when you try to force a particular assay into an automation realm, you can get errors. Not that there is anything wrong with the design of the automation. It’s just not right for the application. The second point is that people assume that automation guarantees a 24/7 operation and that you don’t have to be physically present. They get into the mentality of pressing a button and walking away, and sometimes that leads to failures. Automation does need some sort of human contact. Another disadvantage is that with automation you do need a designated expert. If you need to be fully automated, you need to invest in an expert user, and that’s something people are not prepared for. Some vendors are trying to address this issue, especially with the smaller systems, where new interfaces are easy to use and can be operated without knowing all the complex choices they offer. If you want to make the instrument readily accessible to most people, then work with the vendor to find the main parameters that need to be adjusted. Then your average user can pay attention to only these few parameters and get by 90 percent of the time.
Q: How easy is it to automate cell culture?
A: If you are working with cells as an assay platform, then I highly recommend that the cell culturing be automated, as it is so time-consuming. Automation works perfectly for cell culture, and there are easy ways to monitor if the cells are contaminated or not growing properly. With the new imaging systems you can look at multiple parameters in the cells, such as morphology changes, movement, and more, and it’s relatively easy to do. A key to automating cell culture is to pay attention to potential routes of contamination.
Q: How should people go about evaluating the right vendor for their automation needs?
A: You first have to determine your need to see if you really need automation. If you do, then bring these systems in-house (if feasible) and try them out. People often make the mistake of buying the system first and then finding out that they don’t like it or it cannot do what they want it to do. So map out how easy it is to use the system. Second, find out if it is appropriate for your use. People tend to get systems with all the bells and whistles, when they don’t really need [them]. If you really don’t need a Ferrari, it may make sense to get a Volkswagen.
Q: Do you look at other factors beyond the instrument and the robotics?
A: I look for a company that has a good record of customer service and ongoing support. Cost is certainly a factor when it comes to service, but the relationship with the vendor is very important to me. You may have a great team on-site, but you will still need to involve the service engineers from the vendor on various occasions. The vendor-customer relationship is important, so it pays to do your research. You also have to think about future growth in terms of integrating new equipment. When you are swapping out detectors or other pieces of equipment, they must be able to communicate using a software platform that is flexible.
Q: Is there anything new that is likely to change things in automation going forward?
A: I think there will be little improvements in automation, but frankly I don’t see any quantum leaps, at least on the instrumentation front. In the past few years, the big changes have been more on the software and integration sides. If you have a fully automated system, then there are software packages that can help schedule tasks, minimize downtime, and maximize efficiency. I have seen software platforms that can monitor the log files from your entire system and map out the use and failure times to maximize efficiency. Unfortunately, most people don’t even think about the efficiency gains that they can achieve with automation. The new trend is to monitor the automation system to understand what [it is] really doing and make the changes needed to get the efficiency gains. More changes are coming in the areas of informatics and communications in automation with the use of portable notepads and software that allow control of modular units remotely.