Tour:AIoT Quickstart Guide

Jump to navigation Jump to search

Part 1: Introduction

The AIoT Framework

The AIoT Framework provides good practices and a common language for teams implementing AIoT-enabled products and solutions. The "Why", "What", "Who" and "How" perspectives provide business and technology leaders with a 360 perspective on how to successfully implement AIoT.

AIoT - Why

The "Why" perspective must answer why a company is embarking on the AIoT journey. What is the purpose? And what are the expected AIoT-enabled business outcomes?

AIoT - Why

For AIoT-enabled products, the expected KPI-improvements along the product value chain have to be carefully evaluated and prioritized. Keep in mind that AIoT has the potential to fundamentally change the value chain in the first place, resulting in completely new KPIs as well.

AIoT - What

The "What" perspective must answer what kind of new products, solutions or new features should be realized utilizing AIoT.

AIoT - What

To start with, one has to understand how AI and IoT are to be combined to achieve this - for example by adding intelligence directly to individual assets vs adding intelligence to the behavior of entire fleets of assets ("swarm intelligence").

AIoT - What

Finally, it is important to understand exactly what kind of AI-enabled features can be added to the physical assets. While the details of the specific AI algorithm will be determined later in the project, the expectations on the AI should be validated as early as possible.

AIoT - Who

Understanding and managing the different stakeholders effectively is an important success factor for any project, this is not different in AIoT.

AIoT - Who

The AIoT Framework is providing tools and advise for product managers, project and program managers, development and engineering managers, architects, security and safety specialists, and procurement managers. It also helps to manage internal and external stakeholders.

AIoT - How

The AIoT implementation perspective must look at all the different aspects you will find in any kind of IT project, and combine them with the AI and IoT-specific aspects.

AIoT - How

The "How"-perspective must address the AI and IoT-specific aspects of the business model, leadership and AIoT organization, sourcing, co-innovation, many functional and technical aspects (architecture, data strategy, AIoT DevOps, etc.), and finally legal and compliance.

Part 2: Execution and Delivery

Execution and Delivery

The "How" perspective provides a detailed look at AIoT execution and delivery. This includes combining agile, cloud-based development methods with the usually more safety and security-oriented development methods for asset-centric, embedded hardware and software development.

AI in the Context of IoT

AI in the context of IoT must deal not only with identifying and implementing the best fitting AI algorithm, but also with data strategy (including sensor selection), model design and testing, AI DevOps, packaging and roll-out, and so on.

IoT with AI

The IoT side must usually combine asset-centric (often embedded) hardware and software development, communication services, and cloud (or on-premise) development, enterprise application integration, etc. It also has to integrate the AI-based micro-services into the overall system architecture.

Data Strategy

Both AI and IoT are very data-centric paradigms, so naturally data strategy is playing an important role in an AIoT initiative. AIoT data strategy must include data acquisition, life cycle management, and governance.

Product and Solution Architecture

The requirements from the business model have to be broken down to the functional and technical viewpoints, and eventually to a level of granularity where it can become an actionable backlog for the implementation teams.

AIoT DevOps and Infrastructure

Being able to manage the incremental evolution and continuous improvement of an IT system is addressed by DevOps practices. Extending these DevOps practices to include AI artifacts and IoT-based distributed architectures will be a key success factor for any AIoT project.

Trust and Security

Effectively addressing trust and security will be critical for customer acceptance, as well as protecting the interests of the product manufacturer / solution provider.

Reliability and Resilience

Reliability and resilience will not only be critical for customer acceptance, but also for efficient and effective system operations. This is an essential success factor, and also a main area of investment.

Verification and Validation

The combination of AI and IoT provides quality managers with very specific challenges. Not only is it is difficult to apply established IT mechanisms to the verification and validation of AI algorithms - but in AIoT this also must be addressed in a highly distributed architecture.

Leadership and Organization

The introduction of AI and IoT provides very specific challenges from an organizational perspective, which will have to be addressed by the leadership team.

Sourcing and Procurement

The sourcing and procurement perspective is often neglected by project management, and can quickly become a project`s Achilles heel. The AIoT Framework applies established good practices to the combination of AI and IoT.

Service Operations

Finally, the service operations perspective must include both, the IT Service Management as well as the more asset-centric field service perspective. Different options are discussed.

Case Studies

The AIoT Framework team is currently working on a number of different case studies to validate the framework, and provide reference points.

Part 3: How we work

AIoT Unplugged

The AIoT unplugged sessions is where the work is done. Join us, contribute and learn!

AIoT Training

The AIoT Community is currently developing a training curriculum to make the AIoT Framework more easily accessible. This will be complemented by AIoT Certifications.


The AIoT User Group is the community responsible for developing the AIoT Framework. Join us if you want to find out more!