Optimists at CompTIA predict the number of connected “things” will grow 23.1% annually between 2014 and 2020, at which point 50.1 billion devices are expected to be in the wild. Consultants at Bain have a different view: Their surveys of over 600 executives found that 90% of companies aiming to deploy IoT at some point remain in the planning and proof-of-concept stage. Only 20% expected to “implement solutions at scale” by 2020.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
No matter who’s proven right, there’s a huge business problem hidden in the figures. Why? A majority of IoT devices are distinct, and connecting them can get complicated quickly.
Say you’re driving a sensor-loaded car to your office in the smart city nearby. Alerting the office that you’re coming and signaling the security gate to automatically open as you turn could require a whole combination of devices talking to each other: Bluetooth sensors tracking speed and road conditions, IR sensors signaling relays in roadside lights which connect to the security gate via low-power Sigfox or LoRa. Only a flexible and agnostic platform can connect these devices seamlessly, making sure that data flows as it’s meant to.
IoT sprawl is real
Connectivity is a growing problem in IoT. Why? Devices tend to be easy to deploy. Consider the customer that uses sensors to track energy consumption at corporate headquarters. If utility costs drop 10% as a consequence of optimizing how and when energy is consumed, both the CFO and CIO will be properly motivated to fund new IoT projects. More projects are bound to mean more complexity, because sensors can vary widely in how they collect and transmit data.
Call it IoT sprawl, and it manifests in dangerous ways.
Say the next project is to use sensors for office surveillance, followed by another set of devices for monitoring the health and travel habits of company vehicles. In each case, it’s likely each array of sensors will operate with different security features. They may also have trouble communicating with other “things” or suffer from needing to connect to particularly complex or expensive infrastructure. Flexible and agnostic platforms address the problems associated with IoT sprawl by acting as a universal translator and connector.
A four-part strategy for managing an IoT infrastructure
Creating common ground is a good first step, but the process of generating real value from IoT is also much bigger.
Flexible and agnostic platforms don’t just encourage communication across devices. They also collect, validate and enrich data with analytics before mixing it with other sources and exposing the resulting intelligence to applications for producing business results, whether it’s optimizing energy output for a utility or reducing errors in a manufacturing process.
Check for the presence of these four specific features before deciding on a platform for your IoT infrastructure:
- Support for any device using any connection. Some sensors will connect via 3G cellular. Others will use 4G LTE, Wi-Fi or Bluetooth. Some long-lasting sensors are built for low-power, long-range protocols such as LoRa, SigFox or Ingenu. A comprehensive IoT platform will support them all and the others sure to come.
- Near-unlimited support for IoT connectors. Supporting a communications protocol is one thing. Having a platform that understands and processes data coming from a connected sensor is another, and for that your platform will need built-in support for a distributed messaging system such as Apache Kafka or MQTT.
- Extensive data translation and cleansing. An IoT platform can only be as good as the connections it makes and the data it collects for further analysis. The latter requires a system for translating inputs into a common, well-understood format such as oneM2M so that it can be shared and analyzed.
- Integration with good analytics tools. Well-designed platforms shouldn’t play favorites. In the case of analytics, any platform capable of ingesting machine language should be fair game –whether that’s SAP HANA, tools from Oracle or the HPE Vertica toolkit.
Connecting to a better future
New IoT services are almost always aimed at improving efficiency or driving new revenue. To this end, we’re seeing much more thought going into smart cities and smart energy. We’re also seeing connected cars moving forward while telcos are introducing new SIM-based IoT services to replace declining growth in their landline and data businesses.
In each industry, innovation will be aided by platforms that allow any device, any network and any transmission protocol to interoperate while the resulting data is stored in a common format that can be shared, analyzed and manipulated in real-time.
We all dream of what the internet of things might be one day. First we need the underlying platforms that will allow those dreams to come true.
All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.