Internet of Things (IoT) is essential in bridging the gap between traditional hardware with software systems, business applications with embedded systems, internet protocols & infrastructure with conventional telecom and networking infrastructure. It is making things simpler and manageable.
However, it's necessary to understand the inherent complexity of data exchange among heterogeneous devices, appliances and various embedded sensors etc. before you could leverage its value in improving efficiency or accuracy with minimal human intervention.
There are several IoT implementation projects across the geographies in the area of smart industrial automation, virtual power plants, smart homes/school/hospitals, intelligent transportations or be smart cities.
Often these projects kick start by streaming live feeds of data from biochip transponders or live cameras or in-built sensors or field operation devices etc. with a hope that someone can extract useful data from this and use it for smart decisions etc.
This is where IoT implementation starts to falter. IoT solution team spends days and weeks before finalizing the vendor and tools. During this period an illusion is created that buying the tools and runtime environment will automatically supply the team with "Architecture of the IoT enabled Systems" as well as the Architecture of "System of systems."
In the absence of Architecture, "things" in IoT project starts to resemble a maze of unmanageable hardware components, software systems, data, network elements and service instances which are impossible to disentangle or separate. Also, there are an inherent limitation of current methodologies and frameworks in handling IoT Architecture.
How do we begin and find a better way?
It's quintessential to understanding the six key element types (Data, function, network, role, timing, and rules) that forms the basis of IoT Architecture.
Also, the solution and implementation team must work together to address the following issues:
1. Define, model and integrate IoT Solution Architecture in the context of the Enterprise
for the patient : monitoring heart rates to enable remote health management, support medical follow-up, proper treatment, healthy living
for hospital : timely and accurate medical support based on patients condition, medical diagnostics, clinical lab, decreasing cost of medical service (end to end), increase revenue by support more patients, new pricing models, new service offerings
That means it's not good enough to integrate a wearable solution for a patient, but also mapping how does this help hospital in reducing overall cost in service or increase in number of patients that can be supported in timley manner.
2. Define and model Data Architecture for IoT to cope with a large number of devices which are connected and generate huge data
e.g. creating information for blood pressure, heart rate, pacemakers, wristbands, hearing aids, smart beds, medical follow ups and integration with sensors and bio-medical systems)
3. Define and model Network Architecture for IoT Project
identifying service components for short range wireless such as Li-Fi, QR code, Wifi, ZigBee, medium range wireless (LTE advanced for extended coverage, high throughput and lower latency and long range wireless (LPWAN technologies and protocols such as LoRanWan, NB-IoT, Weightless)
4a. Define and create System (Functional) Architecture for IoT System
e.g key functions such as health monitoring, medical diagnostics, medical follow-ups, disease prevention)
e.g. systems such as Patient information systems, Biomedical acquisition systems, Health data analysis, Digitized healthcare system
4b. Define and create Technology (specification) Architecture for IoT Solution
e.g technology components such as :health monitoring devices (blood pressure, heart rates), implants (pacemakers etc.), electronic wristbands, advanced hearing aids, intelligent sensors, IoMT sensors, RFID electronics, mobile devices,
5. Define and model Quality of Service (QoS such as Reliability, Integrity, Modifiability) for IoT enabled solution
6. Define and model business rules, technical rules and operational rules governing IoT Solution
data collection rules, monitoring rules, emergency notification rules, data capture rules, data transfer rules, data update rules, data analysis rules treatment rules, medical follow-up rules, healthy living rules, clinical rules, lab rules, insurance rules, pricing rules, customer behaviour rules,
7. Identify the critical timing events, timelines, event cycle, event durations, and model them both at the enterprise as well as system components
8. Create process architecture to support automation as devices & sensors are embedded to the manual processes
9. Create necessary IoT solution composites for analysis by linking
linking business strategy elements with business process elements
mapping business process elements with data, function, network, timing, rule and UI elements
connecting system elements (data, function, network, timing, rule and UI elements) to operational elements (data, function, network, timing, rule and UI elements)
10. Use the IoT Architecture elements to create solutions options and arrive at best options using the time tested techniques of simulation, impact analysis etc.
Also, it's essential if you could practice Enterprise Anatomy models to prototype complexity of an IoT Project in terms of technology stacks as well as business drivers. That's the key to ensure traceability and alignment of IoT Services.