Juniper Closes a Gap in Its Connectivity Solution and Moves into Indoor Location

In the beginning of March, Juniper Networks, a provider of products and solutions in enterprise networking and cybersecurity, announced it had entered a definitive agreement to acquire Mist Systems for US$405 million, subject to adjustment. Mist Systems is a company focused on providing wireless connectivity solutions to establishments. Besides offering wireless connectivity hardware and artificial intelligence (AI)-driven network management, Mist also includes an indoor location service that has been regarded as a scalable model that is easy to implement and does not necessarily incur prohibitive CAPEX. With the acquisition, Juniper fills the gap in its end-to-end LAN offering. Previously, the company had to rely on third-party hardware for wireless access points (APs), frequently partnering up with Mist for this purpose. Now that Mist is a part of Juniper, it can reach a wider global market, and Juniper’s portfolio is extended into location technologies. Thus, it is worth evaluating how unique Mist System’s solutions are and whether its business model can thrive in a notoriously challenging market.

The factors that differentiate Mist Systems from most other Bluetooth Low Energy (BLE) indoor location providers can be summarized in three fronts: Wi-Fi/BLE integration, virtual beacons, and AI-based network configuration and management. All three factors make Mist’s solution more easily scalable and quicker in achieving ROI. Furthermore, although some other businesses are known to also work with one or a couple of these same benefits, Mist appears to have found a scalability sweet spot in its combination of the three. Thus, let us now examine these differentials more closely.

The high upfront hardware costs and difficulty in proving ROI have greatly hampered the scaling of indoor location solutions. The main reason for these high upfront costs in many indoor location models has to do with installing hardware. Anchor points, which are the devices used to locate smartphones or tags, are often very expensive to buy and deploy in large spaces. Noticing this, companies like Mist Systems have decided to integrate their indoor locations with the infrastructure of a more primordial service. In this case, Mist includes BLE location capabilities in their Wi-Fi access points. Thus, a Wi-Fi network can be offered a priori to establishments. Indoor location capabilities can be subsequently switched on if desired. This is an example where location is a value-added service. It reduces the CAPEX burden for indoor location by deploying only one infrastructure for two different services. It also helps greatly alleviate OPEX through reduced maintenance, recharging, reconfiguration, and legacy replacement of the infrastructure. Further, it allows the contractor to experiment and verify the benefits of indoor location with the flick of a switch, without incurring risky and expensive pilots. Other companies betting in the Wi-Fi/BLE integration approach include Aruba and Cisco.

The second important factor in Mist’s offering, virtual beacons, leverages the omnidirectional capabilities of Mist’s APs. With the 16-antenna BLE arrays inside the APs, one anchor point is enough to cover a wide area and locate objects and people with 1-3 m accuracy. Therefore, with Mist’s software contractors can program proximity actions like smartphone notifications to happen whenever a person walks into a geofenced region. This is what is called a virtual beacon. For the end user it acts exactly as if they had just walked into range of a physical beacon, without the necessity of actually installing one there. Thus, the system manager can deploy as many virtual beacons as they like with only the original AP infrastructure. Virtual beacons can also be deployed in areas where it would not be easy to install physical beacons. For instance, in a place with a very high ceiling. Virtual beacons naturally reduce the CAPEX associated with buying the extra physical beacons that would be needed otherwise. It also reduces OPEX, since geofenced areas can be readily erased and re-delegated to make room for new campaigns or if the spatial configuration of the establishment changes. For example, the launch of a new exhibition in a museum. Mist is, however, not the only company investigating the potential of virtual beacons. Cisco and Phunware also allow for this in their software offerings.

Last, but not least, is the AI-driven nature of the network management in Mist’s solution. When it comes to location, Mist allows businesses to deploy APs and have the system automatically learn to locate devices through BLE as more phones come into the network. As a result, fingerprinting of the area is not necessary, nor is manual configuration. This removes the burden of initial calibration and simplifies maintenance of the network. This feature is likely to stand out as part of a major trend in the technology world of leveraging the power of Machine Learning to improve a solution. In February 2019, Mist and Kontakt.io announced a partnership to integrate the former’s hardware and network management solution with the latter’s IoT Platform. As made public recently, Kontakt.io’s new IoT Platform leverages what they call Simon AI, a Machine Learning tool to provide business insights based on indoor location data.

Juniper might have agreed to buy Mist mainly for its AI-driven LAN managing and hardware, rather than the location capabilities on top of that. Regardless, Juniper has inherited what could be one of the most promising indoor location models in the consolidating RTLS ecosystem. The idea championed by Mist of location as a value-added service is starting to get more traction in the market not only on top of Wi-Fi infrastructure, but also over smart lighting-as-a-service solutions. When it comes to Mist’s other fronts, for virtual beacons to attract a wider audience, further development on the hardware side will be needed. Allowing for virtual beacons is a simple matter to implement in a software development kit (SDK). However, virtual beacons need to be justified by a considerable reduction in physical AP density. This reduction can only be achieved if the range of current BLE hardware is increased without affecting accuracy. A step in the right direction comes with the standardization of versions 5 and 5.1 of the Bluetooth Core Specification, which improve BLE’s range and implement radio direction finding (RDF). In the short- to medium-term, what is likely to dominate is a mixed solution consisting of both physical and virtual beacon geofencing. At last, removing the need for manual calibration and fingerprinting of an indoor location deployment is a benefit that will be felt mostly by large establishments such as hospitals, where a full-fledged RTLS system can consist of hundreds or thousands of APs. Furthermore, a quick glance at the IoT market nowadays shows that AI-driven platforms are a massive trend and will only tend to increase in coming years. It does seem like Mist is riding all the right waves in the indoor location market. Now, it is in the hands of Juniper to give Mist’s indoor location model the attention it deserves and not let it become diluted. A good first step was deciding to keep Mist’s business and engineering structure intact.