Trends in Industrial Automation

Do I like my job? Yes, I do. Why? Well, it’s interesting. It’s challenging. The company is very innovative, and the brand is associated with technology and quality. I possess a freedom that I can’t imagine existing very many other places. I feel as though I run my own company. If you exclude the monthly and weekly and quarterly and semi annual reports I must generate, I am not actively managed. My director comes from operations with an engineering background. He’s a great manager, in the sense that he ensures the company is functioning, and carries out the requests from senior management. However, he’s not an active manager. It doesn’t know sales, and doesn’t know how to lead. He’s relatively soft spoken. Though he’s very intelligent and thoughtful, he’s not bold, and doesn’t inspire action. All this is very fine. I’m able to maintain autonomy, and still work hard. And I have faith he know’s the pain points, and is doing what he can to come up with solutions.

I plan on applying for the role of group sales manager. The position is open is Chicago and Newark, not my location on the West Coast. However, we need leadership, and I genuinely feel that I possess the hunger, the knowledge, and the ability to inspire a culture change.

We lack leadership. What is leadership? In short, when you have people who know the way, go the way, and show the way. I take full responsibility for my success and failure, and I work around the constraints, rather than criticize them, or complain, or absolve myself of responsibility to achieving whatever necessary because of some excuse, because there is a deficit in support or product or planning. There are givens, and I accept them, and do my best to see the value in what is working, rather than worry about what’s not. And fortunately, we have a lot working for us.

I desperately need to bring in serious sales. What’s stopping me? A few things that immediately come to mind: having an effective “pitch”. Having an effective “value proposition”. Having a thorough understanding of the market, and what products fit where. Following up is a big one. I feel that I’m finding many opportunities for big potential, but I’m having trouble close the deal. Why? Perhaps there hasn’t been enough time? So far I haven’t run into issues of lead times. The biggest challenge is that the value proposition isn’t compelling. World’s smallest is one of our overarching themes. This works, but if its just a nicety, and not a necessity that adds value such as reduce cost or increase performance, then there isn’t an incentive to switch. We’re in need of some unique solutions to solve real problems that cause us to get ahead of the competition. Once we begin solving problems, and there is real value justifying the purchase, then they can begin adopting our other products, which are more or less commodity for most applications.

I could definitely use more sales meetings. Having a LinkedIn account would definitely help me gather contacts and accounts to call on. There’s plenty of companies in my area.

Calling on Biotech or Semiconductor OEM’s is a real challenge for a few reasons. Their redesigns occur every 3-6 years. If you miss a window, you need to catch it the next time around.

Also, our business model, or strategy, is a bit confused.

Our business unit is an industrial automation company. More accurately, we’re a Factory Automation company. That’s all fine. We’re apart of a industrial device division, a $2 billion that sells components more broadly, which includes board level components such as capacitors and sensors and resistors and memory and communication modules, not just factory automation components.

The Factory Automation approach works within a different paradigm. Factory’s are manufacturing consumer goods, and the OEM’s supplying automation machines to factory’s, the ones we sell to, are typically using industrial programmable logical controls, or PLCs, as well as heavy duty sensors with durable enclosures and IP ratings. These systems are relatively simple, running on 24V or analog inputs or outputs. It’s all logic. No data processing, no heavy calculations. Just raw material inputs followed by an automated manufacturing or assembly process, and final goods as outputs.

The machines automating the process are large. Space isn’t really an issue. Ease of maintenance is the biggest concern. Is it easy to troubleshoot and work on this machine? If the product changes, and the process changes, how easy is it to integrate and augment the process? This is especially true for the automotive and appliance and light manufacturing industry. These environments are rough. The processes are simple. There’s not a lot of variation among products, so the systems are relatively straightforward.

However, the Biotech (Lab Automation) or Semiconductor industry seem to be a different beast. The biggest of which is the complexity of the process, size of the final product, and the variation among that product. In both biotech and semiconductor, the size of the final products introduces more quality concerns relating to contamination, precise measurement, and static charge. As a result, the manufacturing environment is a lab setting, with cleanroom requirements.

All this complexity introduces the need for microprocessor controllers, rather than programmable logic controllers. The machines need to be capable of calculating a lot of variation from product to product and process to process. PLCs aren’t designed for this. They need a PC environment to process and transmit data. Unfortunately, PC or microprocessor controls, like Altera or even Raspberry Pi, are custom made for each application, and its a super commodity market with low margins. My company doesn’t make any PC based controls, and it doesn’t have much of an incentive to do so. Beckoff is one of the few. But many bay area engineers prefer to create their own controller, which allows them to retain intellectual property and hedge them against competition. Unfortunately, as they scale, it becomes less profitable to make all the components in house, and soon they’ll need to look for an off the shelf turn key solution.

In the semi-conductor world, manufacturing wafers for microprocessor chips is an incredibly complex process, with many steps, requiring immense tracking and traceability requirements, in a variety of environments ranging from pressurized or vacuum to heat and abrasive chemical. There is a straightforward consensus to the wafer fabrication process, but precision paramount because semiconductors are working in the nanometer scale. However, wafer fabrication is less dynamic and more mechanical in many ways, which makes it a bit more friendly.

In the lab automation world, it’s about micro-fluidics, liquid handling, cell culturing, moving and storing biological samples for diagnostics, analysis, and testing. This world is less mechanical. They don’t care about servo motion precision, or torque feedback, or presence absence. They can get away with stepper motors or linear motors. They’re concerned with moving fluids in the micro-liters, nano-liters, pico-liters. It’s more air or fluid pressure, air or fluid flow, liquid level measurement of micro-plates or test tubes or cuvettes or reagent containers, or temperature. And because we’re dealing with DNA or cells or live specimens, everything is non-contact. The sensor cannot interfere with the process in anyway.

All this presents challenges.

There are many segments within the “Biotech” industry, from microbiology which includes cell culturing and fermentation, to Hematology which involves blood diagnostics, to DNA production to analysis, and these segments penetrate other industries, from pharmaceutical drug discovery to agriculture biology to tissue engineering to immunotherapy. The biotech end users can specialize in a specific process or product for a specific industry. Unless they’re a billion dollar multinational Biotech company, they typically buy third party OEM machines.

For Biotech OEM’s, there are two primary areas of automation. One is Lab Automation instruments, and the other is Lab Automation systems.

Lab Automated Instruments are typically stand alone devices that fit into a typical lab. Sometimes they are large enough to occupy their own space on the floor, but more often than not they fit on a lab bench. These perform instruments perform precise and complex processes at high speed and large volumes that would otherwise be done by a lab technician, which would be slow and prone to human error. The instruments can be moving fluids from one microplate or test tube or container to another, analyzing the specimens, checking the quality of the process like a microplate volumetric scanner, or processing it in some way, such as plate or tube centrifuge. The other thing is that because these instruments are self-contained and sit on a lab bench, size is crucial. This is where my company does well, but not well enough. We produce the world’s smallest sensors, but we’re approaching it from the factory automation paradigm, where IP rating and enclosures are important. The engineers designing biotech instruments are using board level components and sensors, with 3.3 or 5 volts. They need very very small. Small enough to fit into an instrument that fits on a table top, where space is a precious resource. And they need cheap. As an OEM, they’re concerned with margins. These machines are their product, and they need to be competitive. Often all the bells and whistles for factory automation devices are unnecessary and just add cost, so they’ll settle for a less superior but functional sensor at $10-20 rather than $50-150.

Lab Automation Systems are typically larger, and connect multiple instruments together. As a result, the controls systems are processing complex workflow recipes, requiring microprocessor or PC based controllers. The motion aspect is just positioning, so stepper motors suffice for 99% of applications. Often these lab automation systems will incorporate third party instruments, which in a sense takes the automation a step further, and removes the lab technician that would typically move the specimens from instrument to instrument. These systems are designed to be completely automated, with no human intervention. Only the largest Biotech companies such as Roche, Abbott, Hamilton, Seimens Healthineers, and the like design home grown lab automation systems with in house instruments.

So, I say all this to illustrate some challenges.

My company’s current form factor needs to change. There’s a kind of hybridization occurring between factory automation and lab automation that demands a different form factor. One that’s smaller, one that was designed for board level applications and PC controls.

Because we are historically a factory automation company, we don’t have any controls solution, only PLC’s. It may not be worth it to invest or create a microprocessor controls solution. But it also may be an innovative approach to get an edge on the competition. Can we create a microprocessor controller that’s low cost and modular, allowing engineers the flexibility to design their process without building a controller solution from scratch?

We don’t have a motion solution, largely because stepper motors are commodity, low margin items. And they’re replacing servo motors for high speed and positioning applications because of new absolute encoder technology. Do we want to compete in that field? I’m not sure. It’s pretty saturated, and pricing it a race to the bottom.

We do excel at sensors, and we have the technology, we just need to design sensors that fit the lab automation needs. We need to think of fluid processing.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.