A few days ago I had an opportunity to talk to Cathy Zatloukal, CEO of MobileAccess and Steve Lorenc, GM of GE Healthcare's Monitoring Solutions about their announcement right before Christmas – you can read the original post here. Cathy and Steve were very informative and I learned a lot.
First the business relationship. According to Steve Lorenc, this deal provides an enterprise solution to the proliferation of many different RF transmitters, and consequently a lower cost for deploying and managing wireless systems throughout a hospital. The MobileAccess components support the Simple Network Management Protocol (SNMP) and can be managed along with the rest of a hospital's network infrastructure.
Regarding the exclusive between the two companies, GE Healthcare has an exclusive to sell MobileAccess into the health care market. The two companies have been working together for almost a year to develop an intelligent signal conditioning card for WMTS, and do the validation required by GE's 510(k). Antenna design work, electronics, and radio sets were all touched in optimizing GE's WMTS products to support MobileAccess – not an insignificant effort according to Steve.
There is another interesting part to this exclusive deal, namely that MobileAccess is precluded from supporting any other vendor's WMTS. The MobileAccess conditioning cards are designed for specific frequencies and only look at RF, so they don't “see” any protocol. It might be possible to use the WMTS cards designed for GE with other vendor's equipment, provided they're both using the same portion of WMTS. Philips and GE use different portions of WMTS for their telemetry, so any hospital that adopts MobileAccess and wants Philips telemetry will have to deploy their WMTS access points in the ceiling the old fashioned way. This is an important factor in the battle between GE and Philips, and will certainly affect vendors like Spacelabs and Datascope who also use WMTS. In any product strategy that erects switching costs, the vendor who makes the initial sale will have a cost advantage over vendors who try to replace them at a later date (see the book Knowledge Rules in the Connectologist's Reading List). The rest of the medical device industry, however, is moving away from WMTS for a variety of reasons and using the ISM bands (802.11FH and 802.11a/b/g). Vendors using ISM will be able to easily hop on to MobileAccess' infrastructure.
In light of Cisco's recent Clinical Connectivity Suite announcement, I asked how their new relationship might be impacted by Cisco. Both Steve and Cathy mentioned that they've been working closely with Cisco and don't see their relationship as competitive in any way. Of course, if this whole wireless infrastructure market segment takes off, I can't see why Cisco would think they shouldn't own this market as part of their continued dominance in networking overall.
In related news, Steve also mentioned that a new Dash monitor will be running on 802.11b this year, replacing the obsolete (as in discontinued by the OEM) 802.11FH (frequency hopping). He also said that GE would be releasing a cell phone application in the U.S. (that's already released in Europe) for the remote viewing of continuous monitoring waveforms.
Cathy and Steve reported that they've seen broad market interest in an enterprise solution to the current proliferation of wireless point solutions in hospitals. The current sweet spot is new construction and major renovations, but they also see interest – and expect adoption – on a much broader scale. Really, any hospital looking to make a major investment in wireless (and you can't deploy meds admin or an EMR without a WLAN) would probably see an attractive ROI from a wireless infrastructure solution.
During our call, we also talked a good bit about the MobileAccess technology. The two key competitors in the wireless infrastructure space (in health care) are InnerWireless and now MobileAccess. Since I'm more familiar with the InnerWireless solution, we sort of compared and contrasted the two solutions to describe how MobileAccess works. In general, InnerWireless uses passive components in a distributed “leaky coax” configuration. MobileAccess uses active components.
There are two types of RF sources supported by both vendors; single-source transmitters like paging system radios or wireless carriers' micro cells, and distributed access points that are mounted in wiring closets. The MobileAccess head-end takes these central RF sources and “grooms” them for transmission on optical cable for connection to hub units in wiring closets where they're converted back to RF and then filtered and amplified by intelligent signal conditioning cards (there are individual conditioning elements in each closet for each supported RF band or “service”). The RF signals go out from the hubs in wiring closets via broadband coax cable to individual passive radiators that are located for optimal coverage. Wireless LAN access points are connected to the MobileAccess system in the wiring closets (they still need to be connected to your enterprise LAN, rather than the MobileAccess optical RF network). If you don't have any single source head-end transmitters to be supported, you can skip the whole optical RF network thing and just use their wiring closet equipment and antennas. You can read a more detailed architecture overview here.
InnerWireless uses some pretty intense RF engineering to passively support a specific set of RF signals and provide solid coverage throughout a facility. Due to the passive nature of their solution, InnerWireless cannot support just any combination of RF bands, there are some trade-offs. However, the passive nature of their solution is highly reliable; there are no electronics to fail. MobileAccess requires less up front engineering work because they rely on their active electronic components to amplify and shape each RF signal that is used within a facility. According to Cathy, this active approach also allows them to support any combination of frequencies, provided they have designed an intelligent signal conditioning card to support it. (New cards are designed when there is “enough market demand” – my guess is MobileAccess will pay the NRE for broadly adopted technologies and industry standards; while individual vendors probably have the option to pay the NRE for one-off RF applications.) When a facility decides to add a new RF service, say 802.11a which lies outside the 2.4GHz used by 802.11b/g, InnerWireless must “reengineer” their design to accommodate the new band, and may need to redeploy the distributed antenna to ensure good RF coverage. MobileAccess can more easily add new RF signals via additional conditioning “elements” (i.e., cards) at the head-end or in wiring closets. Theoretically, MobileAccess' solution is less reliable because it is made up of electronic components. We did not discuss whether MobileAccess supports redundant configurations with automatic failover when cards or other electronics might fail. Other folks I've talked to suggested that the MobileAccess design and implementation process is somewhat more iterative than the one used by InnerWireless, and may result in going into the ceiling more than once – this is logical on its face, but time will tell whether there is a difference in real life.
Another big difference between MobileAccess and InnerWireless is that MobileAccess use point radiators while InnerWireless uses a distributed antenna. A long distributed antenna snaked through a building cannot support indoor positioning systems (IPS), which is why InnerWireless developed a separate RFID system (more here and here). The point radiators used by MobileAccess act like discreet access point antennas and so they support indoor position systems. So, MobileAccess supports Wi-Fi based RFID systems, and has been working with PanGo, Aeroscout and others. Remember, the requirement here is that MobileAccess has to have a signal conditioning card for the frequency band used by a product. They've got Wi-Fi covered, but don't have cards for vendors like Parco or Radianse.
Wireless coverage is a frequent concern for both users and vendors alike. As wireless devices proliferate though, another factor comes into play: capacity. As Cathy explained it, capacity in the RF domain depends on the amount of energy available at the radiating point. The MobileAccess point radiators, connected to their active conditioning cards can provide higher levels of energy than passive systems.