A long-time reader whose background is clinical suggested I talk to the folks at Meru Networks. I figured it takes a lot to get a clinician excited about IT nuts-and-bolts stuff, so I was happy that Ken Creager, senior director of strategic markets for Meru, agreed to chat. I hear gripes regularly about wireless networks, even with the relatively modest demands placed on them. I was interested to learn more about what’s changed in the time since many hospitals put up their first 802.11b network. Thanks to Ken for the conversation.
Tell me about Meru Networks.
Meru has been in business since 2002. We produce a family of access points and controllers for mission-critical and life-critical environments. The company is headquartered in Sunnyvale, California, with operations in all of North America, Europe, Asia, and R&D in Bangalore, India. We’re not public so we don’t provide financial numbers, but we’re about 280 people, growing at a very rapid pace due to a lot of industry demand. We’re having a great time trying to respond to the needs and requirements of the field.
The lion’s share of our business is in the healthcare and education markets. In healthcare, we solve unique problems as a result of doing a lot of observation in the marketplace, getting assistance from people, and from our participation in HIMSS. We don’t always go in and talk to the technical people.
We look at the nurse as the integrator. If the technology is going to work, it has to be easy to use and functional to a nurse. If a nurse is using a PDA at the bedside, that person doesn’t really care if it’s the applicaton, the unit, or the wireless network if it fails. We work closely with our clients and our partners to make sure we’re very functional for the clinical staff in hospitals.
What’s the penetration of wireless networks in hospitals and how are they being used?
The actual penetration is close to 80%, but let’s clarify. Many of those deployments are first- or second-generation, with fat access points that are difficult to configure and lots of cost. They also tended to have been installed for a single application or department, like something radiology or oncology wanted to put in. It wasn’t pervasive until recently. Most hospitals report that they have some use of wireless, but it’s not pervasive.
What we see happening is an absolute explosion of applications. Go to HIMSS or trade shows and you’ll see applications and devices using wireless as a transport. There was a time when wireless was nice to have, like in the conference room. Today, it’s an integral part of the architecture and an enabler for taking care delivery to the bedside.
We spent a lot of time looking in hospitals and saw this snowball of applications coming at clinicians, but found that networks aren’t pervasive or are limited in their capacity and are failing. Those first implementations may have worked well for an application or two, but with 15 or 20, they are failing. Adoption of devices is not being as well-received as it could have been with a more robust network.
That has given us a window of opportunity to come in and show how our technology is differentiated in the marketplace. We have better coverage and performance and can prioritize traffic to assure application delivery. Let’s say we have a Wi-Fi based phone and we want to make sure that calls get through ahead of someone in the back room who’s Web surfing. We can inspect that traffic, prioritize it, and makes sure it gets through. We have quality of service built into both the upstream and downstream.
A great application of pervasive wireless that we have witnessed first-hand are nurse-type devices like Wi-Fi based phones or Vocera-type badges. You see clinicians walking the hall with those devices. We noticed they stopped walking. They told us it was because they had a good signal and stopped so they wouldn’t lose it. We’re in the mobility business and we asked whether that makes sense. We’ve seen areas where good coverage was marked on the floor with tape. That’s the pervasive element. Is if through the entire facility? Not yet today. We’re getting there.
Common problems in hospitals include dead zones, slowness, and overloaded access points. How does your technology address those problems?
Wireless runs on a series of channels, usually 1, 6, and 11. Access points have different channels and you roam between them, much like when you’re on the cell phone in your car. That inherently causes problems in your end device because it has to continually look to figure out which one of these guys it wants to talk to. At some point, it’s talking to two of them and has to decide how to hand off.
RF planning is required to determine how access points in a general area interfere with each other. Also, as devices move, they have to decide which way to go. If I’m trying to talk to two different access points to determine which is stronger, that’s taking time on the network. Our advantage is that we can put all our access points on a single channel. The end user device sees it as one big network.
There’s no handoff. We make that decision for the end device in our controller. If you’re walking between 15 access points, that entire campus may be on one channel and you’ll never know it’s happening. The advantage is a four to five times performance increase because you’re not asking questions where to go next. Also, it’s seamless between access points. The opportunity to drop a call or device is almost completely negated.
If you think about what’s happening with clinicians walking down the hallway and looking at vital signs on the laptop and they hit a dead zone, they’ve lost information. We take that away because our coverage is more pervasive. We have quality of service upstream and downstream and we guarantee delivery of those packets for critical devices like a patient monitor or voice call. We can assure the delivery of that piece of information.
This all plays into clinical adoption. We’ve seen the reports come out. In the 100 Most Wired, technology today is having a positive impact on health, safety, security, and mortality rates. Much of that’s due to the deployment of technology solving errors at the bedside, medical conflicts, wrong medications, those kinds of things.
Another key thing we find in hospitals is that they’re amass in assets – wheelchairs, infusion devices, phones. The biggest question is “where are they?” COWS and crash carts move to emergency situations, congregate around nursing stations, and then get pushed into the hallway. We can do some locationing with our management software that lets you determine where those devices are.
Because we’re able to do a single-channel architecture of the standard 12 channels, that gives you 11 available. You can stack channels like a stack of pancakes. You can segment your traffic. As an example, you could put voice traffic on Channel 1, data on Channel 6, and telemetry on Channel 11. That increases your capacity on the network and segments them. They can still talk to each other.
Because we don’t have channel conflicts, when you need more coverage or bandwidth, you don’t need more RF planning. You plug in a new access point, it figures out what’s around it, and it becomes part of the community. That’s a low cost of ownership.
Hospitals spend as much upfront with our competitors doing surveys and channel planning as they do on the actual product. We can almost eliminate that. You don’t need as many of our access points to get the same or better coverage as our competitors. The cost of an access point may be equivalent, but you don’t need as many.
When you look at a clinical environment and recognize that a critical care nurse will take 1,000 data points in a shift and there’s five or six of them trying to do something and they congregate, do they have the bandwidth to get their job done? As they move out on the floors, do they have the quality of connection to get their job done?
Also different is that we have an ability to create fairness in the networks. That offers us the ability to do backwards compatibility. You have the b-rated radios that operate at 11 megabits per second. The g-rated ones are at 54 megabits per second. If a guy comes in to your g-network area with a b device, everybody goes down the lowest common denominator. Everybody gets slowed down because of that guy.
We can give all users their full capacity at the same time. We can offer 802.11n megabits, but still allow g and b clients to work on the same network together. In many industries, but especially true in healthcare, devices stay in service for many years. They’re not going to rip out technology to replace the radio cards. That gives us an extensible architecture and investment protection for existing clients.
Describe 802.11n and what impact it will have on healthcare.
It’s the next generation of speed. It will give you six fold the bandwidth of 802.11g. There’s a lot of technical stuff around that, but from an end user perspective, you’re bringing true desktop wired speed to the wireless world.
Most connections to the desktop are 100 megabits. You’re going to have wireless signals that are three times as fast. If you’re building a new facility, do you need to put those wires in place? You can go to the all-wireless enterprise and have speeds faster than that of the wired world.
In healthcare, most of the devices we see are operating very well at b- and g-rated speeds. Ascom has a great g-rated phone purpose built for healthcare with messaging and made for clinicians. On your hip, the display is upside down so you can read it without using your hands. The next generation of phones will have n-rated radios, so you can have more of them out there.
The biggest impact will be in imaging and video. Today’s early generation networks don’t have the capability to take full-motion video or large images. In a shared PACS environment, you might need to look at large images in real time. 802.11 n will allow you to do that.
How important is wireless voice over IP to hospitals?
We’re seeing it as becoming a much bigger element. They view the network as being able to carry everything. We’re seeing dual-mode phones – cellular outside, Wi-Fi based inside. Doctors look like they have Batman utility belts with 15 pagers and devices. You will continue to see an explosive rate of devices coming down and then a convergence period. Blackberry is coming out with a dual-mode device.
Voice is becoming a much bigger element of these networks in healthcare. In many cases, it’s the driver for upgrades. Then, you get into, “What’s the quality of the call? Is it comparable to toll grade? If a bunch of users make calls, is the network degraded?” We have technology that protects the quality of those calls.
What patient care quality issues can result from ineffective wireless architecture?
Time. Let me go back to the nursing station to see what’s happening. If an application is readily available on a tablet PC, laptop, phone, or multi-use device, you’ll save time. The opportunity for errors is reduced. Where you find a low adoption rate of handheld devices and point of care by clinicians, you find higher error rates. Those have an impact on care delivery and quality of care.
If I’m a hospital CIO, why shouldn’t I just buy Cisco like I’ve always been doing?
Cisco has a great product. I used to work for them myself. But this technology is truly differentiated. When you look at a Cisco product, you have no single product in the top five. You’re not really getting best-of-breed in any segment.
We use Cisco products in our demos. We can make their wireless phones work better than they can because our wireless network is so robust. Our technology is extensible and backward-compatible. There are no forklift upgrades. Once you’re set up, you just stick an access point in the ceiling.
CIOs have multiple vendors and multiple levels of code. With us, you have one level of code that runs all controllers and access points. The controller code is broadcast out the access points. You set a corporate policy for HIPAA or JCAHO or whatever is required. Let’s say you allow a certain number of guests, but you have to keep them away from the business office and lab. You set those central policies and the access points come online, assume those rules, and apply them universally across the network however you’ve set it up. Once you’ve set it up, you don’t have to do it again.
We can also suppress rogue access points. Somebody runs down to Best Buy and buys a D-Link box and plugs it into the wall. Suddenly you have a new wireless hotspot with no security policies applied to it. Somebody in the parking lot has access to your network. We have rogue detection. We determine it’s there and don’t let that person come in. We go one step further. Once we recognize that the access point is there and it starts to broadcast, we jam the signal. That keeps devices from taking time away polling the access point. I see that guy broadcasting, I’m going to jam the signal so the end devices never see it and can’t take up bandwidth.
How do you justify the cost of your technology to a hospital that already has a wireless network?
Does your existing wireless network have the capacity to deal with what’s coming? Most tell us no. People with a network in place for 18 to 24 months are having to replace it because of the applications coming. They have to put in an extensible one for the next speed or the technology required.
The advantage we have is that most have already come to the decision that something has to change. We come in and say, “We can solve a lot of these problems with coverage and speed and ROI and save you money as compared to the other vendors, and provide you a better of quality of service.” Our value proposition is strong. Clients are feeling the pain by finding low adoption rate by clinicians on new devices. The end user doesn’t know what’s behind it, it just doesn’t work. We try to build the most robust infrastructure at the lowest cost to make sure those applications work.
Cisco convinced HIMSS to create The Community for Connected Health, which seems to be a thinly disguised Cisco trade group that paid HIMSS for exclusive access to its members. Does that make it even harder to complete against the Goliath?
What’s interesting about that … they did that with HIMSS and had tried to do the same thing with the AMA, who pushed back and made Cisco take down some of their marketing. A week later, Cisco announced their endorsement by AHA. Everyone I’ve talked to on the client side and vendor side says this is an abuse of .org facilities and people. The industry is policing that themselves.
I’ve instructed my team to not even respond to those questions because it’s how Cisco markets today, defensively and protecting their ground. Frankly, I’ve talked to folks like yourself who view that as very offensive, “Cisco has infiltrated HIMSS and I can’t believe HIMSS any more.” I think the industry will self-police that. People who have drunk the Cisco Kool-Aid will buy it no matter what. For those wanting a best-in-class solution, I don’t think them doing that with HIMSS or AHA will influence them in making a purchasing decision.
Wi-Fi companies seem to have had mixed IPO success. Meru was considering IPO this year. What’s the most likely outcome?
We are going through a rapid growth spurt. We just tripled the size of our sales team. There have been some successful IPOs, some not so good, some consolidation. The opportunity for us to move forward and grow this company is excellent. There’s a lot of opportunity out there. We have a disruptive technology. I’m sure the company and its founders and its venture funding would like to see us go out. I’m not privy on whether it’s this year or next or whenever, but when it’s time and the market dynamics are correct, I’m sure we will go out.
Any final thoughts?
Our wireless technology is unique. We’re fully standards-based and we help drive a lot of those standards. We’re innovative in our technology. You’ll find that many if not all of our customers are raving fans of what we do. We have very large hospitals like University of Miami, Wake Forest, and St. Johns. We continue to add and grow in this market almost on a daily basis.
We’re something of a positive disruption. We’re getting a lot of positive write-ups and are getting attacked by people you’ve mentioned [laughs]. When we’ve reached the point we’re being attacked by Cisco, that means we’re a thorn in their side and are disrupting their business. That’s good thing.
The challenge is getting the word out. We’re a small company compared to Cisco. We only do wireless. Customers are benefiting financially. I’m happy with where we’re doing. We’re focusing not only on the IT buyer, but how the products are used by the clinical staff. As we well know, doctors walk in with a great application they found or something they use that they want you to support. We’ll see more and more of that. Having a network that is extensible and easy to add capacity to will have an amazing capacity on the IT staff of hospitals and the budget.