The 5Cs IoT Maturity Model
The concepts and technologies for the Internet of Things (IoT) have been around for awhile, but the business processes and changes stemming from IoT are relatively new. In 2014, John Morada invited us to meet and talk about his ideas for a maturity model. The project was shelved, but his basic concepts are stellar. John gave us permission to continue with this idea.
Update, 2019 November 12: As I’ve become more involved with artificial intelligence efforts again, I’ve realized that our 5Cs model applies to data science as well as IoT. Both types of programs have – or should have – similar goals, to augment understanding and decision making using complex systems that gather, manage and analyze data in new ways.
To grow in IoT maturity, one must recognize that the IoT is an organizational opportunity that encompasses the entire ecosystem. The IoT is not just about the internet, or about information technology (IT), or operational technology (OT), or even bridging IT and OT. The concepts, processes and technologies that have evolved with the IoT, impact every aspect of your products and services.
- interactions with customers or citizens
- design and manufacture
- warranty development and fulfillment
- product quality from field or mine to last use
- customer service
- supply chain
- partner exchanges
- healthcare approaches
- services delivery
- education and training
Here are the basic steps for how an organization should incorporate IoT concepts into their products and services. The business process changes will be considered in further writings and through consulting and advisories.
- Connection - a simple connection pulling data from the device’s sensor-actuator feedback loop(s) and using any wire-line or wireless connection to get that data onto the Internet, an Intranet, to an organization’s Core, that is the data processing facility, or private, public or hybrid among multiple Cloud(s)
- Communication - one step beyond simple connection, wherein one connected thing can recognize, acknowledge and, perhaps, trust a nearby connected thing, person, or intermediate aggregation point, or further up the connection chain – this might be two appliances in a smart home or two cars on a highway or buildings in a campus
- Collaboration - once communication is established among nearby, or not so nearby, connected things, these things might start collaborating, for example, appliance in a smart home might coordinate actions to prepare the home for a returning inhabitant, or connected cars might coordinate about traffic hazards ahead
- Contextualization - from collaboration, connected things will start to add context to each others’ data sets - this may occur at the Edge (Fog), Core and any intermediate aggregation points. Contextual IoT devices will know Why a device in the ecosystem is present as much as How the device will connect. For example, a collaborating set of devices will comprehend and prepare a home environment for a returning teenager versus the parents…a teenager’s wearable will signal the home and launch the appropriate music preference or run a warm shower after soccer practice
- Cognition - from context and evolving data management and analytics, as well as from non-Von-Neuman architectures for cognitive computing, context will lead to cognitive behaviors among the connected things including artificial narrow intelligence (AnI) and general artificial intelligence (GAI)
There are five other “Cs” that permeate each step in maturation of IoT in an organization: Computing, Consuming, Convenience, and Command and Control.
- Computing, as embedded computing or ubiquitous computing, requires an IoT architecture that is distributed and considers the business, operational, system and technology aspects from the sensor package through the chains of interconnections into the core and back again at every intermediate point of data aggregation and decision making. While computing is not a necessary feature of the IoT, Edge computing for processing data for transactions and analytics, as well as managing the data for security, governance, quality and preparation is increasingly important and moving closer to the sensor package.
- Consuming the information and augmented decision making possible from IoT is done by things, customers, citizens, partners, and employees. An important question is the ownership of the data, how the data can be shared, and who may profit from these data sets. I believe that the generator or creator of the data owns the data, and the value comes from the generator-consumer relationship forming sensor analytics ecosystems (SensAE), whether it is one percent improvement providing millions of profit in the Industrial Internet, or creating new lines of business and new services never before considered.
- Convenience is the mainstay that must be part of the value that any IoT initiative brings. The lack of convenience is why past attempts at home automation, as an example, failed. Convenience must also be built into the IoT solution for privacy, transparency, security and building trust through two-way accountability, which is why I also talk of privacy, transparency, security and convenience as being provided, flexibly, in response to situations, culture and regulations.
- Command and (5.) Control have been a part of industrial systems for a vey long time, going back to analog computers and relay-logic PLC (Programmable Logic Controllers) and SCADA (Supervisory Control and Data Acquisition) systems. Current command and control systems often make automatic responses according to pre-programmed rules. The IoT is evolving from rule-based sensor-actuator feedback loops, to pattern recognition and response, to machine learning scoring at the Edge or even in the sensor package, to fully autonomous decisions throughout the ecosystem.
There might be yet another “C”…creepiness as predictive algorithms, pervasive data collection, surveillance and location-based alerts makes one feel as though privacy is lost.
One change that has already happened for many organizations is that the organization’s main “data processing facility” may be a single data center, but is more likely to be distributed among on-premises computers and/or various Cloud services, software defined networks, and platform, infrastructure and software as a service models – the Core; any of which may store, use and contribute to the data sets, analytics, and decisions. The nuances of IoT data and data management and sensor analytic requirements are also being addressed, as well as the grosser data realities of size and speed and distribution of IoT data. Three other “Cs” that impact data usage are Capture, Collect and Contribute. The IoT architecture must consider where and how to capture and collect data, including factors such as sampling rate and various storage points. In addition, the appropriateness and ways in which to contribute IoT data to legacy systems and to broader ecosystems need to adhere to architectural specifications for quality, governance, privacy, transparency, security and convenience.
In my presentations and writings, and in our consulting work, we address the options and specific steps needed to fulfill the organizational opportunity presented by the complex system of systems that is the IoT. We take an anthropological approach to the ecosystem of people and processes through cultural, regulatory, economic, political and environmental factors (CREPEf), and we take a system engineering and architectural approach to the process and technology adaptations. We encourage incorporating the IoT in an Agile fashion, to incrementally adjust to rapidly changing scientific, technological, engineering, artistic and mathematical (STEAM) advances, from academia, governments and industry; as such we have incorporated our 8D Implementation Method with our 5Cs IoT maturity consulting.
- dynamically based on proximity or need
- repetitive data – what needs to be captured
- Wireless (Radio, Optical, Induction)
- Over a Dozen Connection Standards
- regardless of network topology
- act independently when not connected; pick up the conversation when reconnected
- recognize trusted things; reject unknowns
- another dozen or so communication standards or protocols
- sensors sometimes lie; corroborate among similar sensors
- local (sensor-actuator) and intermediate aggregation points and core
- data packaging standards but no true collaboration standards
- collaboration and contextualization both require the nuances of IoT data management to be well understood
- peripheral vision gives context to the central perception
- operational, environmental, duty cycle, maintenance
- nuances of metadata, master data, time-series and location at sensor-actuator package, edge, aggregation points and core
- resolve the event in the context of the human situation, problem, recommendation, scheduling, whatever – without unnecessarily impacting other parts of the infrastructure or environment
- outliers or anomalies – context defines the difference
- there are IoT specific standards for ontologies, semantics and categorization, as well as metadata and master data standards and best practices
- contextualization and collaboration may occur in either order, and even in the same phase of IoT maturity; context may be required for true collaboration to occur
- Machine Learning, Deep Learning, Artificial narrow Intelligence
- scoring at the sensor package, Edge, intermediate aggregation points, Core
- augments human stewardship and use of IoT data and sensor analytics
- can compensate for bad data
- augments direct sensor data for better inferences and decisions
- autonomous decisions as opposed to automating rules
- General Artificial Intelligence could one day arise spontaneously
- There are no standards for cognition…we don't even have a broadly accepted definition
You can listen to a podcast [on iTunes] [or listen to this podcast on libsyn] where I was interviewed to go deeper into IoT concepts, technologies, processes and people skills using the 5Cs maturity model as a framework to build out an example.