Abhijit Jog on Biotech, Batch Automation & ISA-S88 Standards

This episode covers the biotech industry and the push over the last 20 years to automate facilities as much as possible to increase efficiency, and reduce contamination and batch loss. We discuss the implementation of ISA-S88 standards before they even existed, as well as tips for effective alarm management and running efficient reports that prioritize batch quality. We also give a sneak peek into the future of personalized medicine, monitoring hospitals through the cloud, and more.

“People are excited about Ignition and it’s great to see because the development of SCADA systems has been held back by industry for too long. It shouldn’t be a big secret. It shouldn’t be this expensive.”

Episode Transcript:

Lauren: Hello, and welcome to Inductive Conversations. I’m Lauren, subbing in today for Don. We’re here at ICC 2019 and we’re sitting down with Abhijit Jog from Panacea Technologies. Abhijit, thank you so much for joining us today. How have you been enjoying the conference so far?

Abhijit: The conference has been great so far. I’ve learned a lot, and it’s nice to see the enthusiasm people have for Inductive Automation. That is something I think has been missing at other conferences I’ve attended for other vendors.

Lauren: Well that is so nice to hear. We get a lot of excited and enthusiastic people here at the conference so it’s always kind of fun to see all of the different industries collaborating and all of that kind of stuff. So we are really glad that you were able to be here as well.

Can you talk a little bit about Panacea and your role within the company?

Abhijit: Sure. I’m the founder and CEO of Panacea Technologies. Panacea Technologies started as a system integrator in 1996. And since the beginning, we’ve been doing biotech and chemical plants, and we specialize in batch controls. 

So, what we specialize in is doing extremely large and extremely complex batch projects. So it’s not unusual for our projects to have 10,000 I/O for example. And most of these projects are used to make super high-value drugs. So the drugs that we make might sell for billions of dollars.

Lauren: Wow. Now, you’re in a unique position because your industry does work in that biotech and pharma space, but you also have this automation focus. How did you bridge that divide, and what was that process like for you?

Abhijit: Originally I used to work for a company called Life Sciences in the early ‘90s, and Life Sciences was developing some of the first biotech plants in the world. I worked in their automation group, and basically, early on we found out that the biotech industry needed a lot more automation just because the value of the product that was being made was so expensive, that you couldn’t afford to have any operator mistakes or anything of that sort.

As a result, what was not apparent in the first couple of plan designs became pretty obvious in the next few plan designs where people wanted more automated systems. They didn’t want operator intervention. And that industry has just grown and exploded since the early ‘90s.

Lauren: And how did that kind of shift the tone of the whole industry. How has that changed biopharmacy?

Abhijit: What has changed is, earlier, people used to come to us very early on and say ‘Hey, design this plant for me, but I want my operators to have control.’ And lately what we’ve been seeing, is people coming to us and saying, ‘Hey, make it as automated as possible because with automation we can reduce batch loss.’ You can imagine for a biotech manufacturer, losing a batch might mean losing 10 million dollars or something. And it happens. There are operator errors made. And the other issue it the criticality of the product. If you think of a biotech drug that’s made, it’s typically made using something called mammalian cells. And mammalian cells grow at a linear rate compared to bacterial cells that grow at an exponential rate. So all the cells are grown in a serialized environment. They’re transferred through sterilized piping to bigger fermenters, and so on and so forth.

Now if there’s even the slightest possibility of any contamination whatsoever, then you could lose the whole batch because the bacteria cells would just overwhelm the mammalian cells. So as a result, keeping track of all that in an automated system is very critical. That avoids operator errors and that avoids batch loss.

Lauren: Wow. And how has that been received in the biotech space? It sounds like now, people are much more likely to jump in and be excited about it. Has that always been?

Abhijit: No, originally we designed a plant for a biotech company in the bay area in the mid 90’s, and we made an automated system for that. The plant was criticized as being only automated at point. Fast forward ten years later, the biotech manufacturer has found out that this plant is two times more efficient than the next plant in their system, and there is minimal product loss. So I think the people who were criticizing the old automated facilities soon came around to looking at the dollars and saying ‘Wait a minute, this makes sense to spend the money up front rather than later.’

Lauren: Absolutely, you can’t really argue with those numbers can you?

Abhijit: No.

Lauren: Now, within the biotech space, it sounds like you guys focus on batch?

Abhijit: Yeah, we focus on batch automation. And the reason for that is, if you can think of a typical biotech plant, they sort of know what is going to be involved with growing those particular cells, but people really don’t know. And the plant itself might cost upward of 253 million dollars. But by the time the plant gets to production, the product they were designing the plant for might not be competitive anymore. And as a result, what they want is a plant to manufacture multiple products throughout its lifecycle. And the only way to do that is to have batch automation where you can make a product and then switch over to a different product. And as a result of that, they invest a lot of money switch over to batch manufacturing.

Lauren: I see, and so you guys mostly focus on that.

Abhijit: Yeah, we focus on designing the plant to appropriate standards because inevitably you run into situations where they’re trying to make a product that couldn’t be made with the existing standards. So we try to make our standards flexible enough so that the newer product can be made in the plant, while at the same time being able to manufacture products that used to be made in the plant before that.

Lauren: Now I know that you are responsible for one of the largest biotech facilities in the world, that’s ever been created. Can you talk a little bit about that project?

Abhijit: Yeah, the project was originally executed in the mid ‘90s and the tricky part there is that there were no ISA-S88 standards. The standards were in development. There were no standards for electronic batch records. But we wanted our plant to have all of that stuff, so we were designing the plant with those standards in mind and trying to look ahead and figure out what the standards would be. And I’m happy to say that we achieved almost all our goals designing that plant. Whatever we did during the design of that phase essentially made it into the S88 standards and also into the 21 CFR 11 which is the development standards for batch records.

Lauren: Absolutely. And some of the early challenges that you faced in that project, just anticipating the future, how did you even go about that?

Abhijit: Well we kind of had some people who were actually on the S88 committee. And again, the standards are nothing but good practice. So if you can come up with the best practices to put in place, standards are just an amalgamation of those practices.

Lauren: Well, that is a nice way to think about it. And I know that you are part of the group that is helping with 21 CFR Part 11.

Abhijit: Yeah, originally we helped give the comments back to the government on 21 CFR Part 11 and essentially the government entity that was in charge of developing it wasn’t familiar with electronic technologies. So basically they issued a set of standards and then withdrew all the standards and all of the guidelines three or four years later, and then came out with guidelines that actually work with computers instead of just being there on paper.

Lauren: That sounds like it was a little bit necessary. So can you explain 21 CFR Part 11 for people who maybe are unfamiliar with this?

Abhijit: Yeah, Part 11 is basically a way to design reports so that they can be submitted electronically to the government or so that you can have electronic records. And what that means, is when they used to make production batches for pharmaceutical plants before, operators would record information such as ‘Okay, I saw the temperature at this point,’ or ‘I added this product or this raw material to the batch.’ Now what we do is we capture all those actions electronically and all the reports are then forwarded to QA for batch release. So QA can essentially take a look at those electronic records and then approve them to release the batch. Or if the FDA comes in a year or 10 years later, they can look at the electronic records and they know exactly what happened during that batch. So it’s a way to minimize paper and also improve the speed by which these lots can be released.

Lauren: And how has the industry responded as far as you know to this shift?

Abhijit: The industry is still catching up with it because you sort of have to work, for a most efficient 21 CFR 11 system, you have to work backwards. You have to say ‘Okay, what are the things that are important to my product release, and then what other information do I need in that lot?’ And a common example is alarms. So if any alarm happens during the development of running a batch, you basically need to report it. So now, as to what needs to be an alarm, that has to be looked over with a fine-toothed comb because in the early days of automation, people wanted everything alarmed. So now every alarm you need to verify that it doesn’t have an impact on the batch quality. So just a little pruning of alarms at the beginning saves you a lot of time on the tail end of not having to address every alarm.

Lauren: That sounds like a great tip for those working with it. Are there any other tips you recommend for people who are working on their Part 11 compliance?

Abhijit: I think it’s important to get all the groups together to figure out what actually needs to be in the report. And a common example of this has been when you do a sterilization or a cleaning cycle on a piece of equipment, sometimes QA comes along and says, ‘Hey, I want to see the trend that happened during the sterilization to make sure the temperature didn’t fall below 121 degrees or something.’ And our common response to them is, ‘Okay, why do you want to see that? We have a PLC there which is looking at everything every 30 milliseconds so there is no way you can actually look at that, at a trend, and figure that out. And we have a software program that if it does fall below 121 degrees it will flag you with that, right?’ But somebody has to sit with QA and give them the explanation and explain to them that what they’re looking for has already been achieved. And once you do that, then your batch reports become smaller and batch releases become much faster, but it has to be done up front.

Lauren: Interesting. Now what projects are Panacea currently working on? Is there anything interesting you guys are developing right now?

Abhijit: Yeah we have a product out there, we call it the OpenBIO. And essentially what that is, is a fermenter control system that we have developed. And we have developed it using open-source materials, basically. So what that means is, if you go buy a biotech controller from a different vendor, what they do is white-label the parts, and when you white-label the parts, if something goes wrong or some part breaks down, the end-user has no way of getting that part other than through the vendor itself. So in our case we basically tell everybody where the parts are, and they can go look at it. The other technology we have put in there is that all of our biotech controllers have a QR code on it so you can walk into the plant with an iPad, you can scan a QR code, and it will bring up the controls of the biotech fermenter. So there is a lot of innovation that we are putting into this industry.

Lauren: Wow, that is really exciting. What is the relationship like between the open-source community? We have a really positive relationship in the automation community, especially with Ignition. What’s that like in biotech?

Abhijit: It is still developing. And again, it is because it has been very vendor-centric. But what we are trying to move the community towards is, you know, to share more essentially because the end-user companies are in the business of making medicines, they’re not in the business of competing on their automation platforms.

Lauren: Now I’m curious where you see your company headed next? What are you big goals for the future?

Abhijit: Yeah, we are actually looking to develop products for the personalized medicine space. I’m sure you guys have heard of this, where you take the T-cells from a person who has cancer or whatever, and basically grown the T-cells that fight the cancer, and then inject it back into the person right? So that is personalized medicine. You’re developing medicine that works only for one person.

So what we are trying to do is develop products for it and we think that these products will go into hospitals rather than into the manufacturing plants. And hospitals basically don’t have the depth of talent to run these systems. So what we are hoping to do is, we’re connecting these systems to the cloud so we can monitor them from a remote location, and essentially the hospital staff can run it the same way they would run the washing machine for example.

Lauren: Wow. How is the medical industry thinking about this?

Abhijit: It’s still in the early stages.

Lauren: Yeah.

Abhijit: Things are still developing, but personalized medicine is coming because the rates of cures for certain types of cancers are just off the charts, and it’s basically training your immune system to go fight cancer cells. So it is coming.

Lauren: You’re automating your own body system, right?

Abhijit: Pretty much, you are automating the cure to your own body system.

Lauren: I like it, that’s very compelling. And as a whole, where do you see the biotech space headed toward this personal medicine? Or other thoughts?

Abhijit: Yeah, the biotech industry itself, just based on the biotech drugs, is growing quite significantly. And what is going on is, there are biosimilars now. And biosimilars are similar to genetic medicine that you take for small molecular drugs. And as a result of the volume, the number of biotech drugs simply increasing exponentially every year. And the market is also increasing for these. As these drugs become cheaper, people from less developed countries can now start to afford these drugs. Earlier these drugs cost too much to manufacture, and as a result, it was not possible to use them in less developed countries. But now, with the volumes that you’re starting to see, things are starting to change.

Lauren: That’s really interesting. I’m curious where Ignition fits into all of this. How do you guys leverage Ignition?

Abhijit:  Well currently we leverage Ignition by using its abilities as a SCADA platform to scale up manufacturing pretty fast. We are hoping that we can deploy Ignition at some point in the future in the cloud and monitor all of our machines through Ignition which exists in the cloud. So we are working on development for these things we use where you can go and scan a barcode and have the fermentor pop up on the screen. So there’s a lot of places where we see Ignition being used. Also for the reporting aspect of things, we use Ignition for generating smarter reports for lot release and things of that sort.

Lauren: That’s exciting to hear. Well, I want to thank you so much for coming on the podcast. I’m curious if you have any final thoughts or things you’d like to share with our audience?

Abhijit: I think it was personally very strange for me to come here to ICC and see so much enthusiasm during this conference. People are excited about Ignition and I think it’s great to see because the development of SCADA systems has been held back by industry for too long. It shouldn’t be a big secret, it shouldn’t be this expensive. And I hope to see the same values go down into the actual controllers themselves because the controllers are overpriced. And now we are starting to see controllers from Ignition partners such as Opto 22 which are more realistically priced.

Lauren: Well Abhijit, thank you so much for coming on the podcast today. I’m really grateful that you took the time to talk to us. I hope you have a great rest of your ICC.

Abhijit: You’re welcome. Thank you.