Industry reports promise major benefits to OEMs that add IoT capabilities into their end-products and who adopt data-driven business practices:
- Improve operational efficiency & reduce operational costs
- reduce device servicing costs by optimizing device performance through sensor monitoring of critical components (battery, motors etc)
- use of predictive maintenance services to reduce device downtime & outages
- device operational data can be used to make “real-time” decisions
- demand forecasting and service request prediction
- Generate new revenue sources through new service offerings,
- Build better products through long-term direct measurement and analysis of product operations and customers use,
- Improve user experience: the most responsive brand, improve CSAT and build customer loyalty
- Improve agility: immediate insights from new software features deployed with customers.
This is achieved through collection of device data from sensors embedded in the OEM device itself or in the environment in which the device operates or both. At regular intervals sensor data is
communicated to an aggregator or into the cloud using any number of communicationn networks such as ethernet, WiFi, Bluetooth, Zigbee, 6LoPAN, cellular etc and a number of open network protocols such
as MQTT, CoAP or proprietary. This data communication can happen using a heartbeat as short as a millisecond or as long as hours. The data can be stored in the cloud (or in edge aggregators) and
enterprise applications can be built to display, report on and perform simple or deep analytics which yield the BI required by the OEM's business.
The goal of the Internet of Things is therefore to acquire data from various embedded systems and impart analytical processes on that data to improve performance, efficiency, and business
outcomes.
If you are looking to add IoT capability to your existing products or you're developing new products with IoT capability and you're looking for a customized IoT solution,
RTSS can help.
Looks simple but, in practice, selecting an IoT platform, initiating an IoT project, deploying & running IoT products is highly complex involving many moving parts. IoT projects involve:
- Operational technolgies or OT (the edge devices and aggregators an OEM ships to its customers),
- traditional Information technologies or IT like databases, cloud servers & storage,
- Communications technolgies for edge device connections and cloud connection,
- Security technologies to provide end-to-end device & data security,
- Application technologies such as data visualization/reporting/analytics that yield BI value.
The runtime platforms you develop & deploy with your products are an important part of your decision making process. Low-end, simple IoT devices, such as may be found in a home (smart
lighbulbs, temperature controls, entry systems etc), may need a simple standalone runtime based on a low-end MCU with MPU with a communication stack; but, as device complexity increases you're sure
to move up to an MPU with MMU which will require a runtime based on an RTOS and with high-end products that require rich-media support (e.g. video & audio) or high-end UI with 3D graphics or a
web-browser and multiple communication stacks, you should be looking at runtimes based on Embedded Linux or Android.
IoT device architectures now typically involve multiple CPU cores so additional platform capability and additional development tool capability is required for core communication/management and
application development & debug on these multiple core architectures. Additional software tools are needed to develop multicore IoT applications.
Device user-interfaces are also changing from the traditional embedded device interface; device users now expect a device UI which is like their smartphone. Embedded UI Software development
tools are needed to develop your custom device UI.
So, it is likely that you will deploy IoT-enabled products with different runtimes ranging from standalone, simple-to-complex RTOS or Embedded Linux and your device management platform will have
to be able to manage these different device configurations.
Security of embedded software has typically not been a major concern as historically embedded devices have operated as standalonbe devices performing a specific function. IoT-enabled devices are
connected to a network via some form of communication such as Bluetooth or WiFi which increases attack surfaces and therefore the risk of being hacked. Device OEMs need to consider adding software
security layers into their end-products to minimize the likelihood of device operations being compromised or critical data stolen.
Having an aggregator at the customer premises gives you the ability to centrally manage your IoT products for that site and also have local device-to-device communication and co-ordination.
Real-time decisions can be made about what data is passed into cloud storage for deep analytics to reduce data traffuc flow and cloud storage costs. Real-time analytics can also be provided in the
aggregator to provide real-time decision making and improved responses to incidents.
The aggregator is optional, however: another supported architecture is where you have a single device at the customer premises and in this configuration the cloud drivers, which normaly reside
inside the aggregator box.