In this video presentation, Divya van Mahajan, Head of Digital Networks at Philips, provides an overview of the company’s recent experience successfully scaling operations to meet increased demand during the COVID-19 pandemic. Philips is a leading health technology company focused on improving people’s lives and enabling better outcomes across the health continuum.
When the COVID-19 pandemic began, Philips needed to rapidly ramp up production of key equipment, such as ventilators required by frontline care workers in intensive care units. This explosion of demand meant Philips had to aggressively increase its manufacturing to meet a fivefold increase, while also transitioning as many as 52,000 employees to remote working. During this intensive period, its IT operations team had to adopt a strict zero-downtime strategy. Mahajan describes how the team leveraged ThousandEyes End User Monitoring and Enterprise Agents to support this effort.
Follow Along with the Transcript
Divya van Mahajan:
Philips is a company which deeply believes in health care and, actually, at the core of Philips is innovation. And we focus on delivering innovation that matters to you. Here you can see a product that is much in the news today: a respirator. Which helps people breathe and especially people who are infected with COVID-19. Getting this kind of technology into the hands of health care providers becomes very important as you start trying to get on top of this crisis, as well.
Divya van Mahajan:
A bit more about Philips. We strive to make the world healthier and more sustainable through innovation. Our key goal is to improve the lives of 3 billion people a year by 2030. And I think this resonates very well with me because I've been with Philips for over 18 years, and I've seen a big transformation from looking at lighting and looking at semiconductors towards focusing more on personal health and health care. Today our businesses are in 3 main domains: diagnostic and treatment businesses where we look at MRIs. We look at C.T. scans to provide insight for providers on what's going wrong with you. We provide connected health, connected care, and health informatics, which allows people in the ICU to monitor you very carefully. And lastly, we have a personal health business that takes care of you at home. Looking at things like mother and child care. Looking at all the health care, the toothbrush lines, personal care and domestic appliances for our kitchens and food appliances.
Divya van Mahajan:
So you start looking at this you start seeing a lot of touchpoints with people. Okay. Which are very relevant today, as well. It is interesting to me to see. Where do I fit into this picture? When I was pulled out from architecture, from the cloud, I was asked, please come in and transform a network. And I started taking the same approach that I took in the cloud towards a network, as well. Can we use more software-defined? And we transform from a network and telecom department into a digital network, where we rely on SD-WAN, unified communication, cloud gateways and remote access. Very different from a few years back where we were just a more traditional network department. Our current team is a global team and we manage over 250 sites across the world for Philips.
Divya van Mahajan:
When COVID-19 came, it came a bit earlier for us than March. March is when, let's say, the rest of the world woke up and said this. But all the way early in January, we got a crisis call from China saying, hey, something is raring up. We need to start being prepared for everyone being shut down in Wuhan. And that was a wake-up moment for us. Okay, the transformation of how we work has been quite tremendous and all of you have seen this and experienced it firsthand.
Divya van Mahajan:
What it meant for us was people start asking to work from home, as well as talk about scaling up the VPN, high increase in the use of video and Microsoft Teams, virtual desktops. You've seen people stopping traveling, yet downloads and events like this take place. So we need to support these on the existing infrastructure and also ramp up our digital online channels as we have a big increase in our online shopping, as people stop going to retail shops. Finally, and probably the most important is there was also a call to action. We provide healthcare equipment to providers and a lot of countries, a lot of providers needed to rapidly ramp up the ICUs and ensure that they have the correct equipment for front care workers. And this meant a translation into Philips increasing its manufacturing tremendously to meet the demand.
Divya van Mahajan:
Below that, you see some of the technologies that we tried to use from ThousandEyes to help us in this sport. We looked at End User Agents, Enterprise Agents to support a monitoring performance and ensuring that Philips can continue operating during COVID-19 crisis. I'm going to look at two cases up here. One is work from home. What we noticed is in the short span of a few days, we had to start ramping up to prepare VPN concurrent capacity to go from 20,000 users to 50,000-52,000 users at the same time.
Divya van Mahajan:
Now, at the same time, people also started ramping up the use of Microsoft Teams, whereas previously most people were quite happy with chat and audio. The lack of personal connection meant a lot of people started looking at video calling and we saw an increase from 6% of normal call rates to 30% of calls now using video. And I think you guys have experienced a similar thing in your personal day to day use. On my side, when I look at my children. I see them using Microsoft Team, I see them using Zoom. And I'm quite proud to tell them, you know, that's what I do in my day to day work. They say, “Oh, yeah, that's great.”
Divya van Mahajan:
So when we started helping people shift to work from home, we found things were not quite as easy as they'd hoped. A lot of people are just not used to working from home. They're used to working from the office. And we start seeing that there's a different way of application performance or people perceive performance between onsite on-premise offsite and using virtual desktop services.
Divya van Mahajan:
So we had to rapidly figure out a way that we can measure performance from all these locations without sending anyone into the offices. And that is where ThousandEyes was really helpful. We installed the Agents in all these locations, and we could compare performance across different places. So when we had a challenge from our research stuff saying work, very rarely work from home, coming to us saying our research stuff is really hampered by performance. Can you help us?
Divya van Mahajan:
So the researchers are based in Cambridge, Bangalore, Haifa. They have a research cluster, scientific clusters in all of these locations. And they were trying to replicate what they did in the office from home. And they were complaining about poor performance, but they weren't really sure what was causing the problem. Since this was new, they didn't have a handle on what was a necro part, etc. and traditional ways of diagnosis, like using a tracer etc. was kind of out of the question because the geographic's better than researchers know we could fix the Internet, we could fix a VPN, we could fix it VDI, or you could even fix the application.
Divya van Mahajan:
The question is, we needed insight. What is the best way to achieve this goal? And this is one of the challenges. I think Sonya also touches on it. What you can't see, you can't measure and what you can measure, you can derive any actionable insight from.
Divya van Mahajan:
To illustrate what is actually meant and you can see that global spread up here, I'm taking the research team from Israel. Yet people in Israel working from home offsite, yet people working onsite. And they were saying our experience, working from home is very different than working from the office. They had to path for them to come into the office, either coming through the VPN that's in the center or coming from their VDI, which is off to the right. They also were accessing SaaS applications, as well as on-premise applications, which are located in Israel EMEA as well as the US.
Divya van Mahajan:
With this complex, mix of applications. Okay, how did we get to a more objective way of figuring out what needs to be fixed? The first thing was to measure. So what you see here is we started using the Endpoint Agents and the Monitoring Agents to figure out what's the latency between a person sitting at home in Israel to the VPN, to the VDI. And what is the latency that we can measure all the way to end-to-end to the applications?
Divya van Mahajan:
Pretty soon what we noticed is it really depends on what application they are trying to use. When you looked at the first application in the US, you could see a performance when we are going through VDI was, on the whole, very similar to what they would be getting from home. Same thing from the TFS Netherlands application. The big rub was really the on-premise applications for which were hosted in the Israel office itself. And there we had the double whammy. First, they had 80 milliseconds to go to the VPN or VDI, and then that VDI had to come all the way back to Israel, which added another 80 milliseconds to the latency time.
Divya van Mahajan:
So at the end, after looking at this analysis, the answer was very obvious. Make it easier for them to connect directly. So we can call a local gateway in the Israel office and allowed the Israel researchers to access it directly. But this actionable insight, I think, is what I think of as the real value of having monitoring consistently and easily scalable across the network.
Divya van Mahajan:
But taking the same case back to a manufacturing site. Okay, what you see up here is how Philips helps our health care providers. We have diagnostic equipment. So you see a machine for looking at your lungs, etc. with MRI. You have ventilators, which are used in the health care ICU units. You have the monitoring for them. And these needed to get to our providers quickly as people started getting nervous about the number of people who would be coming into the intensive care units. So solutions have physiological monitoring in the ICU, and they also had during ventilation, clearance therapy in the ICUs.
Divya van Mahajan:
When people started putting more orders in, we had to ramp up production. And the head of operations for Philips told I.T., I can't afford any downtime. Our customers are relying on us. We need to ensure that we get these things to them. And I have 24/7. I need to deliver 5x production. I can't give you any downtime at all. And how did we meet this challenge? I think here the challenge was trying to be ahead of the curve of manufacturing, trying to see, Okay, are we going to have an outage in the next half an hour or not? Can we give them a quick ramp-up? We had residency already built-in. So what we really needed to have is monitoring.
Divya van Mahajan:
And that's where ThousandEyes steps in, we were asked to ensure that there'll be 0 downtime. And our key to production was they could manage a short outage of 5 to 10 minutes as long as critical ERP, MES applications were still accessible. Since locations are spread all over the world. Here again, ThousandsEyes came to our rescue. So what we used is ThousandsEyes at each of these locations. We started monitoring each of these applications and could quickly turn around and provide an application performance dashboard.
Divya van Mahajan:
And you've seen examples from Sonia as well as some steep on how this kind of a high-level overview gives you a very good insight into where the bottlenecks and the problems are. Here, for instance, the red ones are indicating that access from the Singapore VPN towards the second application is showing a very high latency. So we can quickly jump in and say, is that normal? Do we expect it or not? And the jumping in is when the second you quickly finding the bottlenecks helps us.
Divya van Mahajan:
With the end-to-end view, latency graphs as well as our ability to time travel, we can compare the before and after situations in a crisis, as well as alerting. And this is really helped give the confidence to our manufacturing teams that, yes, they had the data and they'll be on top of that. And this is where I really like to be. And I like to thank the team from Orange as well as ThousandEyes for helping Philips achieve its goals. And in turn, I'm quite proud that Philips is able to get this equipment back to people who can help patients. Thank you.