Showing all posts tagged healthcare:

It’s WLPC Time Again

The WLAN Pros Conference is truly a unique experience that I look forward to all year long. Throughout the year we are inundated with vendor marketing material and embroiled in competition. WLPC is a few days where we can come together as individuals, educate each other, build the community and challenge each other to be better at our craft. This year’s conference will be in sunny Phoenix, AZ. Read more about it here. If you’ve never been before and you have an interest in Wi-Fi I urge you to make plans to attend. It is a great opportunity to network and learn from others in the field.

This environment provides a great opportunity to get up and speak about something you are passionate about. The mix of longer presentations and ten talks allow for a lot of variety and depth of topics. This year I’ve selected a topic on Healthcare wireless as my main presentation topic and then will use a Ten Talk slot to provide a sneak peak into my Bluetooth World presentation that I will be giving in March at Levi’s Stadium.

Designing Wireless Networks for Clinical Communications

Healthcare presents one of the most challenging wireless environments in today's networking world. The unique blend of critical network applications and expectation of high speed ubiquitous wireless access for everyone is challenge enough and then numerous devices are layered on top. Clinical communications are critical to providing a high quality of care and has become an especially challenging environment to plan for. This post is intended to offer some guidance in designing these networks.

The Emergence of the Smartphone as a Clinical Communications Tool

Smartphones are joining the healthcare scene at increasing rates, companies such as Voalte, Mobile Heartbeat, PatientSafe and Vocera are bringing new features and functionality to market and are transforming communications at the point of care. These devices are typically either Apple iPhones or the Motorola MC40, however plenty of other variations exist. Each of these phones have numerous differences in how they behave. This differences vary from when they roam to how they handle packet loss, etc.

Access Point Transmit Power

In preparing to design for a clinical communications network a desired endpoint should be known. In almost all cases, Smartphones tend to have lower transmit power than what most admins are used to. As a result, we are designing wireless networks with transmit power of 10-12dBm rather than 14-17dBm as many legacy networks were built. This reduction in access point transmit power drives up the number of access points required to cover a facility by 25-50% depending on construction.

Data Rates

Disable lower data rates to reduce network overhead and functional cell size.

Access Point Placement

Fast roaming is critical to the performance of Voice over WiFi and for Smartphones this typically means leveraging 802.11r and 802.11k. Understanding how these protocols work and their impact on roaming is essential for success of any network being designed to support clinical communications. As a wireless engineer tasked with this design, the goal is to create small, clearly delineated cells with enough overlap to facilitate the roaming behavior of these mobile devices. If designed poorly, 802.11k can be a detriment to device roaming. Some general guidelines to follow:

  • Access points should be mounted in patient rooms and out of hallways whenever possible
  • Leverage interior service rooms to cover longer hallways--clean storage, food prep, case management offices, etc.
  • If you must place an AP in a hallway
    • consider planning to use short cross unit hallways rather than the long hallways wherever possible
    • consider using alcoves to your advantage to reduce the spread of the RF signal
  • Leverage known RF obstructions to help create clean roaming conditions that favor 802.11k
  • Overlap may need to be as much as 20% due to roaming algorithms in the smartphones
  • Pay attention to the location of patient bathrooms, facilities where these rooms are in the front of the patient room (near hallway) offer far more challenges than those where it is in the back of the room
  • Stagger APs between floors such that they are not vertically stacked on each other

Voice SSID

Configure for a single band whenever possible - you'll find that some vendors are still only comfortable with 2.4GHz. From experience this can work, but is not without issues either. As a general rule, I recommend using AppRF to view the applications using the SSID and prioritize them properly. Smartphones are always talking via multiple apps on multiple ports and this should be accounted for.

All Apps Are Not Created Equal

Certain mobile communications apps are simply not ready for the demands of a healthcare environment. Take the time to understand exactly how these apps are being used, on multiple occasions I've seen perceived "dropped" calls as an app issue rather than anything to do with the wireless network itself.

Test, Test, Test

This is still a relatively new application for Voice over WiFi and it will require effort to get it right. Extensive testing is typically needed to get these deployments 100% dialed in. Tuning from AP placements to transmit power tweaks should be expected to some degree.

Hyperlocation, A New Era of Location Services

I had the opportunity this week to visit Cisco Systems with the the Wireless Field Day 8 crew this past week to discuss the updates to the Cisco location services roadmap. This space has been a passion of mine for a number of years and to see major advances in the last 12-18 months has really gotten me energized again about the possibilities that are out there.

Cisco provided a recap of how we got to where we are today and I wanted to do the same to ensure that the concepts here are captured properly.

Location services use the triangulation concept to position a device within some probability of accuracy. In the example below, the object being tracked is roughly equidistant from the access points in the upper and left access points and a bit further from the right access point. As you may already know, under most conditions the best accuracy that can be expected is 3-5 meters and there is no certainty that the tracked client will be in a specific room as it could be on the other side of a wall, etc. This has lead to minimal adoption of location services as often this information simply isn't good enough to be useful. It's also important to note as Cisco mentioned in their discussion that for a packet to be seen by all three APs in this scenario, it must be a broadcast packet. On many devices, broadcast packets are becoming more uncommon while associated to an access point. This means that the refresh rate of this location information is slowing down, decreasing its usefulness further.

One of Cisco's recent innovations to assist this refresh rate issue is known as FastLocate. This uses data packets to improve the rate at which we see packets. Since data packets only exist between the client and the AP that it is associated (green arrow in the FastLocate diagram), the other APs have to go off-channel briefly to hear these packets as well. This process improves the refresh rate from approximately 1 frame per 60 seconds to 1 frame per 10 seconds. It's important to note that while this does not improve the actual location measurements, it does provide more data points which allows for better overall resolution and causes less "jumping" in which clients appear to teleport from one location to another.

Cisco's Hyperlocation Module (HALO)
The HALO module was the highlight of the Cisco presentation for me as I had not seen one of these up close yet. The HALO module leverages a 32 element (16 2.4GHz and 32 dual-band 2.4/5GHz) phased array antenna that sits on a collar surrounding the access point and attaches via a special connector on the new wireless security module. This array is comprised of directional antennas which enables far greater location accuracy and is truly the first real leap in wireless location accuracy in quite some time. Since each of these antennas are facing different directions and are all slightly located apart from each other, it is possible for the APs to calculate Angle of Arrival (AoA) in the wireless signal. This combined with the traditional triangulation allows for very specific accuracy down to approximately 1 meter, which is competitive with BLE capabilities. Further testing will be required to see if these claims hold up and if we can see actual room level (or better yet bed, etc.) accuracy, but I believe strongly that this will be the case in most situations. Currently this limited to associated clients only, but there is a near term roadmap for products that beacon on regular intervals such as RTLS tags. Tracking unassociated devices is of lesser priority right now, but is on the road map.

Halo to Usher in a New Era of Location Services for Healthcare
Enabling an infrastructure with serious location capabilities isn't cheap at roughly $1000 per access point (MSRP), so the use cases need to be substantial. HALO is a premium engagement and location offering that will not be adopted by the masses, however those who can leverage it effectively stand to offer tremendous value to their customers, patients, stakeholders, etc. Hospital environments to be the single most valuable environment for HALO, especially after RTLS tags are added to the mix. This environment is highly mobile and ripe for disruptive location technologies that can automate work flow and enable meaningful interactions and engagement. Hospitals today leverage multiple location systems that facilitate portions of work flow, however very infrequently are they integrated together well enough to gain additional actionable insights. There are a number of point products addressing nurse rounding, locating wheelchairs and beds, theft prevention (babies and assets), etc. Some of the "bleeding edge" use cases include playing a loved one's voice over IP speakers to coax an elderly patient back into bed or having a display cheer up a child who is battling cancer when they near the screen. In theory, this HALO module is enabling a world in which we know where every associated wireless device is down to a matter of feet, without necessitating RTLS tags. This opens the data floodgates for actionable intelligence at unprecedented levels.

Work Flow Optimization
RTLS vendors have been selling their solutions for years to address these issues and while there are clear ROIs in most cases, it still doesn't get adopted due to significant cost and/or limitations for a single purpose system. HALO offers the ability to gain insight into numerous inefficiencies within a hospital and can improve any investment already made in tags and RTLS systems. More importantly a truly location-enabled work flow could facilitate meaningful interruption ensuring that nurses are only getting notifications when absolutely necessary and routing other requests to someone who is better positioned to assist a patient quickly. These use cases have been well defined over the years and HALO is an enabler to making these a reality.

Throughput Analytics
Integrated with Prime and the MSE, HALO can provide improved throughput analytics for the waiting rooms. I'd expect this to get down to individual care arcs in which an organization could track a single patient progressing through their visit and providing trends. This would provide feedback into the work flow optimization process and enable further analysis of the changes made to the work flow processes.

Patient Engagement
There has been a lot of buzz around "mobile engagement" lately and as a professional in this field I can attest that there certainly are many options and the jury is out as to which is the best one at this point. I am a huge proponent of the Low-Energy Bluetooth solutions out there, however HALO may be the first "Wi-Fi only" offering that offers the ability to truly engage with BLE-like location capabilities. Organizations evaluating these solutions need to account for the care and feeding of a BLE-based solution as like many of the RTLS offerings there is a battery powered infrastructure that has to be maintained.

Turning Low-Energy Bluetooth (BLE) Upside Down
One of the major limitations to the Low-Energy Bluetooth solutions is that they are almost exactly backwards from RTLS solutions in that mobile assets are required to interact with fixed (usually) assets. This typically means an app on a Smartphone (carried by a mobile user) is the typical mode of engagement and this is an active process. A few use cases of cropped up recently in which a BLE-beacon is strapped to a patient to monitor them as they move through an environment. To accomplish this the interaction points along the way require a tablet mounted to the wall to pick up on the patient's tag and initiate the engagement work flow. This reverse BLE-solution is somewhat innovative and solves a specific problem, however we need a world in which mobile assets are interacting with other assets on demand and relative proximity to each other should be irrelevant as long as I have access to the back end data. I believe that HALO is positioned to do exactly this as it evolves and in doing so will allow the use of Wi-Fi based geofencing capabilities indoor, reducing or possibly even eliminating the need for traditional fixed beacons. At the very least this will provide us new deployment options and can be combined with a BLE infrastructure to offer infinite interaction capabilities.

Internet of Things
The items above all point towards enabling the Internet of Things (IoT) and allowing real time interactions between devices to the betterment of the healthcare environment. An example of this could be the lifecycle of an infusion pump. These devices follow a predictable cycle in which they move from being clean through being attached to a patient and ultimately wind up having to be cleaned again before being released back to the care areas. In this scenario, alarms could be triggered by a device that has not been put in a maintenance mode and is following an unexpected path, possibly avoiding a dirty pump being brought to a patient, etc.

Using HALO
Effectively using HALO requires that the APs are mounted to the ceiling grid as they need clear line of sight. Density is expected to be at least one AP every 50' with each covering roughly 2500 square feet. Environments using smartphones, etc. will typically see smaller square footages per AP due to the requirements of those smartphones, so the density of these modules will be higher.

Installations are expected to take longer due to specific needs around mounting locations. X,Y,Z coordinates need to be exact as well as the HALO module rotational orientation and alignment with antenna 0 of the 32 element array. These positions must be set in Prime 3.0 to provide proper location capabilities.

HALO Requirements
Cisco 3600/3700 AP (and future access points)
Cisco Wireless LAN Controller running 8.1 MR3 (8.1.123.x)
Cisco Prime 3.0
Cisco Mobility Services Engine
Cisco Connected Mobile Experience (CMX) 10.2.1