Skip to Content

How to Run “Internet of Things” Projects

When the Machines outnumber Humans they will need an Internet of their own. It will be known as the “Industrial Internet”. This could well be the line from a future Terminator movie, but the reality is that Industrial Internet is already here. With billions of devices and communication modules at a low cost and technologies maturing at a faster pace the Internet of Things or IoT on the backbone of the  Industrial Internet is set to grow in leaps and bounds.

IOT Trends.PNG

Fig 1: Trends impacting Internet of Things

There are the  Network companies who are interested in building Infrastructure for the Industrial Internet, Device manufacturers strengthening the security of their devices and business models being rewritten to unlock the opportunities provided by the Internet of things. The opportunities provided by the convergence of several technologies are immense for both businesses and consumers.

One of the earliest consumer use cases of the Internet of things was the Nike Plus running app on the Ipod. A Microchip combined with a piezoelectric sensor device placed in the shoes of the runner would send the running data once per second to a listening transreceiver in the then Ipod which would get the data on the speed & calories, etc. This model has evolved with the smartphone era to provide enhanced experiences to the customers with deep analytics on the run as well as commercial applications using the data. 

Nike Model 2.PNG

Fig 2: Layers of the Nike Plus Service

While this is a simpler analogy to understand the tech stack of IoT in the case of Industrial companies, there are thousands of equipment which are spawning terabytes of data every day that need to be collected, transferred, stored, processed and made accessible to the users in a sensible manner. This information can then be harnessed and put to use to develop new customer offerings, offer preventive maintenance, predictive analytics, optimize efficiency and increase user engagement thereby creating new business models. In the industrial world the architecture looks somewhat like this.

Iot Architecture.PNG

Fig 3: IoT Technology Stack

While the hardware manufacturers and telecom networks dominate the first two layers of the stack SAP has a major role to play in the rest of the stack with its technology, database and business solutions to extract the value.

To handle to an ever growing number of device types, network protocols, communication methods, deployments, data processing, monitoring, alert management the following functions are needed.



Device Modeling

Abstract view of devices(Attributes ,data & Control functions)

Relationship to Real world entities

Device Hierarchies

Device Instantiation

Device Registration

Device Configuration( Remote & Mass Configuration)

Device Lifecycle Management

Device Communication

Secure Communication with devices

Synchronous & Asynchronous

Streaming & Batch

Device Control

Remote Device Control

Device Health Checks

Executes commands described in device model

Device Data Processing

End to End data flows from the edges to the center and back

Modeling with Design time tools

Instantiation, monitoring and auditing at run time

Device Alert Management

Dealing with Alerts from devices(Locally & Centrally)

Manages states of alerts as defined in the device modeling component

Alert Notification and alert query processing

Operator Console Services

Services for Real-time operator consoles


Tables/views in Hana

The business applications of the Internet of things are far and wide into the intersections of different industries. A Machine could easily turn into a data hose which could then be used to make processes more efficient and green while providing better customer experiences. Some of the examples are as below.

Real-time Asset Monitoring

Consumables Management

Connected Vending Machines

Remote Diagnostics and Updates

Marketing – Cross Selling

Remote Health Monitoring

Warranty and Recall Management

Usage Based billing

Smart Home energy management

Fleet management

In-Store Media Channels

Machine Monitoring

Pay as you use

Greener Waste Management

Real-time tracking

Smart Parking

Meter Reading

Driving Behavior analysis

Electric Vehicle Charging

Maintenance Field Services

Consumer Applications

While the possibilities are endless with what one can do with Internet of things there are not many methodologies for choosing which business area to IoTize and how IoT projects should be implemented.

For companies considering Internet of things applications the journey should start with the customer as the consumer of the value derived from the Internet of things. Does setting up IoT make your offering more efficient, more intuitive, more green. In this age where the distinction between Service and product companies are blurring Internet of things gives an excellent opportunity to build “Customer Solutions” rather than products or services in a silo. Customer solutions would include the hardware, software, backend infrastructure, user interface, pricing and business model. A Service development methodology for product services is adapted to IoT projects as shown below.


Fig 4: Proposed Project Methodology for IoT Projects

Idea Management: An Idea day with a cross functional team for Brainstorming different offerings could kick start the process.

One way could be to look at the Machine Data being currently generated and probe ways to make that Data more useful for people running the business or the consumers. Example: If the Business has a variety of field machines which log data such as usage, run time, diagnostics, loads, frequency. The available data can be analyzed and presented to users with intuitive dashboards and leading to more profitable service mechanisms and usage based billing models.

The other way it to look the overall process and choose the right points at which machine data should be captured to add more value to the process. Ex: If Field service personnel are making frequent trips for regular maintenance without adding much value it would be helpful to have a set of sensors capturing key performance data to provide proactive and on demand service.

Requirement Analysis: The requirements to meet the idea need to be analyzed from Market, Customer and Internal capability assessments. The most important piece would be to identify the Information Consumer. Is this information being consumed by the Employee, Customer of the actual service being provided or the Service Consumer. The value derived for each entity is different from one another. The Customer might be interested in the total cost implications and visibility into the nature of billing, whereas the service consumer might be interested in more intuitive, easy to use and modern UI. The Employee on the other hand would be interested in getting the processes more efficient and Integrations to other service entities to make the tasks easier.

Service Design: The next phase consists of conceptualization. This is made up of Service Definition(Scope of Services, Technical description) , the organizational concept and the Marketing concept( 4Ps= Product, Price, Place, Promotion). The main goal of this service is to elaborate the service in more detail, setup the organization for provisioning and deployment of the resources.

Service Simulation: This stage is the prototyping stage where the service is sketched out along with all the actors and interfaces in it. The test should include simulation of the concept for value and Simulation of the architecture. One of the key simulations would be a simulation of the Interaction with all the IoT elements, including  the user touch-points and business models. The main goal of the simulation is to envision the end state and evaluate whether it is providing the intended value.

Service Implementation & Test: This stage is the Development stage where the service is developed in all the layers of the IoT stack. The Product modeling is to identify all the “things” that are generating the data need to be captured and abstracted using Device modeling, Device instantiation and communication. The service composition involves composition of single services to build specific applications. The process modeling of the device instantiation, registration,retirement, security related features, privacy related features also need to be considered carefully during the service implementation stage. At the Business model layer if there are commercial implications to the user like Usage based billing or maintenance plan triggers at specific data points, then those services need to be built as part of he process modeling. Both the product model and service model needs to be tested at a Unit level and at the overall architecture level for Integration.

Launch: The service is then rolled out and in most cases is a phased rollout to iron out any issues in the service layers giving a chance to improve the user experience for the wider users. The IoT project can be concluded with a feedback from the customer and the Information consumer. This feedback loop is critical for a IoT project to identify areas of improvement and managing the interoperability efficiency of different systems.

The utility of the IoT deployment needs to be constantly measured using KPIs such as Availability, Incidents in the application, Active and Passive usage and interaction from users( Ex: Access to data reports from users), commercial impact including RoI of the efficiency gains or the commercial impact of the IoT enabled Billing models. 


The Internet of Things: A survey, Luigi Atzori a, , Antonio Iera b, , Giacomo Morabito c,* aDIEE, University of Cagliari, Italy,  bUniversity ‘‘Mediterranea” of Reggio Calabria, Italy cUniversity of Catania, Italy

Service development for product services: a maturity model and a field research , Thomas Burger1 , Walter Ganz1, Giuditta Pezzotta2, Mario Rapaccini3, Nicola Saccani4

1 Fraunhofer IAO, Stuttgart (Germany), 2 Università di Bergamo (Italy), 3 Università di Firenze (Italy), 4 Università di Brescia (Italy)


Be the first to leave a comment
You must be Logged on to comment or reply to a post.