A new article published by me on the Campana & Schott Web site.
Artificial intelligence sounds like expensive high-performance computers and complex algorithms. But with the right data and technologies, it’s really easy to get started. Especially if you start with simple use cases and platforms.
As production manager, Sabrina wants to prevent the frequent accidents in the aisles of her factory. To do this, she has cameras hung up that show if an object is in the way. An automatic system is to warn the respective shift supervisor with an alarm sound and a red light in order to avoid personnel expenses. The task sounds simple, but as is often the case, the devil is in the details. Because it has to clarify:
- Which technologies are suitable and which are available via IT and cover the needs for this use case?
- How should the solution be integrated to derive added value from the AI use case and link the data to the warning system?
- What data is needed for the AI system to distinguish objects in the way from mobile transport vehicles or people?
The right technology for the use case
To find the right technology for their use case, companies should evaluate application requirements based on two categories: available expertise and customizing options. Four levels of complexity are presented below to help select the right technologies.
- Complexity Level 1 – Integrated Platform Components
- Complexity Level 2 – Out of the Box Services
- Complexity Level 3 – Low Code/No Code self-learned machine learning models
- Complexity Level 4 – Custom Data Pipelines with Machine Learning Models
Integrate the technology
However, anyone dealing with AI technology must not ignore the important aspect of interfaces. In fact, these are the key to success. After all, technology and data alone do not generate added value. They must be used for the right use cases and integrated into the organization with the appropriate processes.
Therefore, it must be clarified how the technologies can be linked with applications and processes so that the overall process runs cleanly. For example, the data generated by AI usually has to be used by other systems, such as CRM or marketing solutions. Azure often lends itself to this as a platform, as AI solutions can be smoothly integrated with Teams, Office or other Microsoft applications.
In the above example, this means linking object recognition with the warning system for the shift supervisor. Here, the highest warning level calculated by the AI must result in turning on the red light and activating a warning tone. To achieve this, the compatibility of the two systems must be checked.
… read more within the full article (German): Technologie – Motor für die Super-Power KI | Campana & Schott (campana-schott.com)
