How do you get a well-respected, 80-year old maker of laboratory instruments to embrace the cloud and the Internet of Things — stat? That’s the task facing Wido Menhardt, VP of Clinical Informatics at Beckman Coulter, headquartered in Orange County, California and part of the $17 billion Danaher Corporation.
Menhardt spoke with InnoLead about getting the organization to take the cloud seriously; how the company uses employee hackathons to explore new technologies, languages, and platforms; misconceptions about technology adoption in emerging markets like India and Africa; and letting the voice of the customer drive innovation.
What We Do
We are in this business of testing blood and other fluids for a large variety of markers, chemistry, like lipid panels, or immunochemistry, like viruses, or hematology, like your white blood count, or microbiology, looking at bacteria.
We do about 10 billion tests a year…With that, we probably touch over half a billion people each year. For these are just approximations, but one thing that’s said in this industry is that 70 percent of medical decisions are based, in part, on blood tests.
It’s a huge impact that our instruments, and the results they produce, have. It accounts for only about something like two percent of health care costs. It’s a huge cost–benefit ratio.
That’s the same for IVD [in vitro diagnostics.] We have a full range of instruments that a clinical laboratory would need. Clinical laboratories are typically either in the hospital, that’s for in–patients, typically.
Then there are commercial labs — like a Quest or something — where typically when you’re doing your annual check-up, that’s where the samples go to and where you get the results from. We can equip both with our instruments.
Now in addition to just having the instruments, we don’t want the lab techs to go around with pencil and paper and taking note of what the instruments need. In addition to that, we provide sophisticated software that helps orchestrate this.
…The way it works is that somewhere, somebody creates an order. Maybe in the hospital, maybe your doctor. It gets entered into some software system, and that then arrives at our software, which then makes sure that all the instruments know which instrument has to do which test for which blood sample for this particular patient.
That’s orchestrating things on the incoming side. Then when the results are available from the all the instruments, then our software goes into, if you wish, intelligence mode, where it does many things, but it does two very important things for the lab.
How We’re Organized
We are organized in business units that focus on certain things. There is a business unit that’s for amino acids and chemistry, and a business units for hematology. The business unit that I’m in focuses on informatics and other automation solutions. That’s my little box.
But with next-generation software, I’m also responsible for the basic software platform for our next-generation instruments. Those next-generation instruments, we obviously want to have user interfaces and usability that’s comparable to our Cloud solutions.
We want those instruments to be connected to our Internet of Things solution and send really valuable data there so they can service better but on the other hand so they support next-generation analytics, whether it’s clinical or predictive analytics of failures, etc.
Connected Instruments and the Cloud
….[Currently, our] software is installed on PCs in the lab or servers in the lab, etc.
We’ve made a strategic choice about a couple years ago, to take our solutions into the cloud. All of our instruments report data via Internet of Things into the cloud. We probably have more than 10,000 instruments connected to the cloud that send enormous amounts of data every day.
On the back end, if we put all of the data into the cloud, provide several advances to our customers. One is that if the solutions are in the cloud, they can access it anytime, anywhere. Whereas right now they have to go through a PC in the lab. Now they can look at it in the cloud. They may be at different locations in the lab or in the hospital. Maybe ultimately at home.
All of those obviously have to be managed from a regulatory perspective. Another advantage is, if you have all of the data in the cloud, you can also do data analytics. Look at patterns or inferences from all that data that you’re getting.
Right now, the data is in a PC or in a server in one lab. If the hospital has many labs, then aggregating that and ultimately even looking at comparing data from different labs. For example, if different labs could compare the results for similar tests with similar chemistry, then they could actually have some sort of a peer comparison.
Finally, if you have them in the cloud, that’s operationally much easier for us and the customer. We can much more easily install and support the software. That’s a benefit to us but also benefit to our customers.
[But for us,] software is relatively new, cloud, and Internet of Things is even more new.
…A couple of points. One is that within the company, talking about using the Internet to support a laboratory, the first hurdle that you hit is that people don’t think that Internet is reliable enough, Internet connectivity is reliable enough for a laboratory.
Somehow, they feel that in this domain it’s all different, even though in reality in their personal lives and even their professional lives, they use the Internet for their banking, for their shopping, for entertainment, and really indeed, the uptime of the Internet for those purposes is highly satisfactory, I would say.
How to Think About Emerging Markets
…Another thing that’s interesting is that sometimes people will then say, “Yeah, maybe in the US. But in emerging markets perhaps not.” Emerging markets are very important to us, China is a huge growth market for us, and further down the road, India and Africa.
People think that, “Oh, but if you’re somewhere in rural India you won’t have Internet connectivity,” but the reality is that mobile Internet penetration in emerging markets is huge, in particular wireless. You’ll find people with smartphones anywhere in Africa, anywhere in India, and in western China.
There are solutions that people are putting in place for Internet connectivity, even in emerging markets, maybe even in some cases more so than in developed markets. The other thing is that people in emerging markets are very resourceful.
In my previous role, I did a bunch of work in healthcare solutions for India and for Africa. In Africa, for example, I’m four hours away from the nearest hospital, which also hardly deserves the name and a local health clinic which mainly catered to pregnant women and vaccinations and those kind of things.
We had put in place a solution for ultrasound for pregnant women so they would know whether they would have a risky delivery.
But we didn’t have skilled ultrasound people everywhere in rural areas. We, in some of these healthcare centers, we put in ultrasound machines, tell somebody how to run the ultrasound probe over the belly of the pregnant woman. But the interpretation would have to happen in that four-hour-distant hospital, and there was certainly no wireless Internet in that healthcare center.
What they did is they took a very long pole, like 30, 40 feet long. On the top of that they mounted a 3G antenna. With that, they were able to send the images to that remote hospital. They’re just very resourceful. It doesn’t have to be perfect, but it works.
How Our Group Is Different from IT
…Another change process is IT. In a company like ours there is a corporate IT department and that obviously runs financial systems and our email and our websites and our general ERP systems, etc.
In a company that’s focused on instruments and chemistry, and I find this interesting because as a software guy I come into this company. Traditionally, people’s intuition when it comes to software is to go and call corporate IT, even if it’s about software solutions that we sell to our customers.
Corporate IT is focused on delivering better service at lower cost every year. They’re trying to use their costs to offshore things, etc. Better service at lower cost. That’s their focus.
Our focus, on the other hand, is providing more value to our customers, expanding our footprint, driving the top line, driving the revenue, more solutions, better value for our customers. It’s a very different focus than corporate IT. Neither is bad. Each has its own focus for a different need.
But crystallizing that within the company and explaining that if we develop and support products for our customers, we have to use one sort of information technology strategy. Whereas if we support our internal users with email and ERP systems, it’s a different information technology strategy, and they don’t align too well.
Regulatory Affairs and Risk
Another is that [regulatory affairs] is something that is very, very important in our company, because our products are highly regulated. It has to go through high regulatory bars for approval. Then for surveillance. Periodic audits, etc. That’s part of the DNA of the company. People are very focused on complying with the regulations.
That means a lot of documents, a lot of approval cycles. These things take a long time. But on the other hand, some of the software technologies that we are trying to push out are either not regulated at all — some of the operational software technologies, for example. If we provide customers with a dashboard of how much testing they do and which instruments at which site, that’s not regulated.
Other software technologies that we have are of relatively low risk and low concern from a regulatory perspective, so we could for those move in principle much faster with much more frequent updates, maybe even as fast as once a month, which in this industry is very, very fast.
Of course, if you look at Facebook or Google, they do it several times a day. But even getting to a month, in an industry, where regulatory methodologies and habits have been entrenched, is difficult.
Distinguishing between what is important to have a lot of documentation and very diligent approval cycles, on the one hand, and then other things that can move very quickly, is difficult. Making sure that we don’t brush everything within the same brush within the company is a big change for us, even for our internal organizations and systems.
Security is Like an Onion, Not an M&M
Another change management topic is security and privacy, because that’s always a topic that comes up when people talk about the cloud. Two things there. One is that the traditional IT approach is to create impenetrable firewalls, and often lots of passwords and virtual private networks, etc. Try to have a perimeter around the organization that nobody can get in, and that way we’re safe.
But with a strategy that takes us into the cloud with lots of interconnectivity from our instruments, via Internet of Things, into our back office, into the cloud, and then to customers, it’s no longer simple to have a wall around you.
We’ve changed our security strategy to be rather than having an M&M, where you have a hard shell and then once you’re through the hard shell you’re in the middle and everything is soft, instead of that we’re using a strategy that’s more similar to an onion, where every onion peel, we have a security strategy for that. Every element in our solution, whether it’s a disk or a network or a user, everything has its own security mechanisms built in.
When I talked to customers, initially there seemed to be a lot of concerns about taking healthcare solutions into the cloud. Customers are concerned about security, privacy, etc., but I decided to do voice of the customer with the hospital CIOs, to see what they think about it.
As it turns out, their main concern is dealing with companies that don’t know what they’re doing, that don’t take security seriously. There have been, as you probably know, lots of breaches in the healthcare space. …But if somebody actually does know what they’re doing, has a good strategy, gets certifications for that from a third party, then the household CIOs are quite happy with the cloud solutions and they’re quite commonplace.
This is an important finding and an important way to communicate this kind of change internally in the company as well, because internally people also hear about security and privacy as a concern and maybe cloud is what we should be doing as a company. But as it turns out, it’s consistent with what our customers want as well.
How We Use Hackathons
One more thing is change management within R&D. My primary job responsibility is R&D and strategy. In R&D, we have extremely high retention. We are a company that’s a good place to work and many of our R&D staff have been with us for a very long time, and enjoy what they’re doing. They come to work every day helping to improve the healthcare of every person on the face of the earth.
But going to the cloud requires many new technologies and tools, new languages, new networking technologies. Privacy and security is also part of that, but in particular on the web there’s always new stuff that makes it easier to develop solutions and connect them and use mobile applications to link it with that, etc.
Some of the teams, in particular in software, may not have been exposed to those, because primarily they’ve been working on software that runs instruments, often embedded, but certainly on-premise, on the instruments. Moving people who’ve been in that space for a long time to help them to appreciate and learn these technologies is sometimes a challenge.
What we’ve done [is organize] hackathons inside the company. We say, “We’re going to take this concept and we’ll develop a cloud-based solution for that. What technology are we going to use for that?”
…We invite five or six teams of two or three people into a room for a week every day, and these teams, we ask them to come with their preferred technology towards doing things, [whether it is] Python or React or whatever.
Then in the room every day we say, “OK, today we want to solve the following problem. I don’t know, user logging, or displaying 10,000 points from a particular assay on the screen, etc.” During that day, each of the teams goes away and hacks it up … and at the end of the day they compare the solution. How does it work? What were the problems they had? Look at some source code, etc. Then towards the end of the week there’s an emerging consensus of what sort of technology to use. It’s awesome, because it’s fun, people are engaged, people feel that they had a say in what we’re going to do.
At the same time, we’ve solved the business problem as to what is the platform we’re experts on.
Don’t Try to Do Everything
…We thought that some things could be done exclusively from the cloud, and by getting feedback from experienced people that work in the field, we had to accept that some things cannot be done in the cloud, at least for now.
We stepped back and said, “OK, maybe for certain situations we need still one, little, on-premise box that does things, for example for speed of communication between devices or maybe as a backup, should the cloud go down.”
There are certain things where we sat back and said, “OK, we’re not going to crack this nut,” sometimes because we thought we might have one conviction, but, for those particular cases, we don’t have the evidence to prove that it all works so it’s better to take it one step, release that to the market, and then see how that actually works, and take it from there.
The Voice of the Customer Can Drive Innovation
The mantra in many companies like ours is, [let the] voice of the customer drive your innovation.
…Conversations with the customers are extremely important. …I have one beautiful, beautiful story that I just love.
I was in Australia at a customer visit. In the meeting were both the head of pathology, so that’s the doctor, who is responsible for the results and all that, and the quality. Also in the room was the lab director. This is the person that manages all the lab techs that operate instruments and also the software.
The conversation went on for some time with the pathologist, on high-level strategic issues and clinical and quality-related issues. Then at some point, after half an hour, the lab director piped up and she says, “My employees, they are on their smartphones every day, and they use all these apps. They’re all in the Cloud.
“Then when they want to start to operate medical devices,” and this is not only lab instruments, it’s anything now, “So when they want to operate medical devices, they have to go back into the last century.”
That was just so powerful because with that, I go back to our organization, internally, and I tell them, “You have to remember…” Actually, at the recent leadership meeting I had a slide where I showed them a day in my life. We start maybe before breakfast or at breakfast.
I look at my smartphone, look at my emails and stuff and then I move on to a laptop perhaps for meetings and video conference. Then I move to a two, three-screen desktop to do some hardcore work with PowerPoint and Excel.
Then towards the evening maybe I’m on my iPad to do some email again or social media. Then at night I have a projection TV so I pull down my screen, I project Netflix or Hulu. Some of the applications that I use, like Facebook or YouTube or email, lives on all of these screens. You have that same life, correct?
I told them, “This is my life, a day in my life. Who has a similar life?” Everybody in the room raised their hand and said, “Yeah, my life is like that as well.”
Then I told them, “Guess what? Your customers have the same life. They have the same expectation. They have the expectation that our software solutions are just as convenient to use and as flexible, when it comes to devices and screens and locations as your personal life.”