Massive IoT Creates Massive Opportunity: How Vendors Ensure Delivery of Device Management Services Across Large-Scale Deployments

Massive Internet of Things (IoT) is both a large-scale deployment of IoT devices and one that is resource constrained in terms of computing power and battery power. Massive IoT deployments are strategically important to IoT companies, like platform vendors and connectivity suppliers because of massive IoT’s scalability, which can create large and recurrent revenue streams. Device agents for the massive IoT can be divided into two groups, the first grouping consists of Short-Range Wireless (SRW) devices which requires a gateway to connect to the cloud. SRWs do not require embedded software, or “device agents” on the SRW device, hence an agentless architecture. However, there is still a requirement for embedded agents on the gateway itself. The second category are IoT devices which do not require a gateway backhaul because the device has cellular (such as Low-Power Wide-Access (LPWA)-LTE) modules built into them. These are increasingly Lightweight Machine-to-Machine (LwM2M)-compliant (support for native device agents) and so can support IoT device management services which are based on LwM2M.

An agentless architecture means there is no customized device agent from the device management vendor deployed on the endpoint device (a tracker, sensor, detector, or meter), because the device is an SRW endpoint using Bluetooth, Zigbee, or Wi-Fi technologies that require a gateway backhaul instead of communicating directly with cloud services. However, the market is at a crossroad due to rollouts of LPWA networks and device Original Equipment Manufacturers (OEMs) are integrating LwM2M agents natively. These module devices are also called LwM2M-compliant devices, and through NB-IoT and Cat-M can provide long-distance connectivity as opposed to SRW. Integrating LPWA-LTE technology into LwM2M-compliant devices means connectivity is increasingly embedded on endpoint devices.

Key vertical segments impacted by rollout of LPWA-LTE and procurement of LPWA-LTE modules include telematics units, asset tracking devices, and metering applications. For these application segments, it is common to use direct cellular connections to the cloud (Cat-1, Cat-M, and NB1). These endpoint devices do require device agents as embedded software so that they can connect to cloud services. In such cases, the trend toward open standards and pre-built LwM2M-compliant modules is accelerating the rollout of massive IoT services. Other vertical segments still require a gateway backhaul, which is typical for home monitoring, home security, and condition-based monitoring. In these application segments, gateways are useful for two reasons. Firstly, gateways provide SRW devices in a Local Area Network (LAN) environment with a Wide Area Network (WAN) connection to the internet. Secondly, gateways are higher compute devices which can provide analytics capabilities at the edge. On the endpoint device (SRW device like a temperature sensor for condition-based monitoring), the analytics capabilities are limited, but at the gateway, those capabilities increase substantially.

The strategy of proprietary software which led to vendor lock-in is increasingly being challenged. Multiple factors, including the rollout of LPWA networks and the rise of LwM2M, have diminished the power of vendors that sought to use proprietary device agents, at least in the massive IoT domain. Strategic decisions by suppliers on vendor lock-in was designed to create difficulty for customers to substitute one supplier with another supplier. Implications of a competitive market environment notwithstanding, generating stickiness rather than lock-ins is a potential win-win compromise both for vendors that seek to maintain recurrent revenue streams and customers that seek to control costs, as well as maintain a degree of autonomy when choosing who their device management vendor may be. For an IoT deployment leveraging agentless architecture on endpoints, routing connections to the cloud via a gateway means the solution is to deploy LwM2M device agents on the gateway itself, which can aggregate data from multiple endpoint nodes and transmit them to the cloud for data processing. For long-distance endpoint devices operating without gateways, hardware OEMs providing pre-integrated LwM2M clients is an ideal solution, however pre-integrated clients are not a cure-all. Some module OEMs which support LwM2M don’t allow firmware-level access and thus device enrollment to a LwM2M server cannot be done by default. The lack of access for a LwM2M server to the device means that a custom embedded software is required to onboard the device onto the device management server. For massive IoT, the computing environment is constrained resource and so LwM2M with Constrained Application Protocol (CoAP) as the underlying protocol is more suitable. LwM2M makes application layer device management services an open standard when compared to other protocols which did not have a standardized application layer. Previously, the lack of open standards had created space for vendor lock-in to emerge.