The days when devices did nothing more than follow user commands are over. Artificial intelligence has significantly expanded their capabilities, giving devices the ability to make decisions.
A robot vacuum recognizes obstacles. A smartwatch analyzes your sleep and physical activity. A security system can now distinguish ordinary movement from a potential threat without user involvement. What is more, modern devices can learn from real-world situations and improve their performance over time.
As AI transforms products, it also changes the way those products are created.
Agile, Lean Startup, Design Thinking, and Jobs-to-be-Done, which are the four most popular approaches to product creation and development, emerged before artificial intelligence began controlling physical devices. This means that traditional methodologies are no longer sufficient.
Today, teams must address far more questions than before: which sensors to install, what data to collect, how AI should respond to different situations, whether the device has sufficient capabilities, and how the system will continue learning after launch. And this list is far from exhaustive.
AI-powered products require new development approaches primarily because all their components are closely interconnected, and the final quality of the device depends on how effectively those components work together.
Imagine that the wrong camera, radar, or another sensor is selected during product development. The collected data may then be incomplete or inaccurate. As a result, the artificial intelligence may fail to assess the situation correctly and make the wrong decision.
At the same time, AI capabilities also depend on the device’s processor, memory, battery, and other technical specifications. Laboratory testing alone is not sufficient for such a product. It must also be tested in the real world, with all its unpredictable situations and unforeseen circumstances.
The AI & Physical Devices Product Framework was developed in response to these challenges.
It serves as a roadmap for teams developing AI-powered physical devices.
It is not a step-by-step product development process, but rather a methodology for making product decisions. Its primary purpose is to help teams understand which critical decisions must be made, when those decisions should be made, and how each one will influence every decision that follows.
This intellectual product does not replace Agile, Design Thinking, or other traditional approaches. Instead, it complements them, particularly in cases where AI effectively controls the behavior of a physical device.
What can established methodologies do? They help teams identify the right idea, organize software development, or improve team processes. The AI & Physical Devices Product Framework has a different purpose: it helps align decisions concerning sensors, AI behavior, the technical capabilities of the device, and its continued learning after launch.
It does not introduce new sensors or algorithms. Instead, it brings separate disciplines together within a unified system to demonstrate how a decision made in one area will affect all the others.
The AI & Physical Devices Product Framework was built on practical experience developing real AI-powered products. The projects differed, but the questions that arose during their development were largely the same.
Practical work revealed which approaches were the most effective and made it possible not only to refine them, but also to document and organize them into a unified system.
“The AI & Physical Devices Product Framework reflects my years of experience addressing product management challenges at the intersection of artificial intelligence, sensing technologies, and physical devices,” says Illia Bondariev, the developer of the framework.
He notes that physical devices are becoming increasingly intelligent. As a result, the challenges faced by their developers, and addressed by the framework, are becoming not only more novel, but also more important.
Let us summarize the key factors behind the success of AI-powered products. The power of the AI itself and the technical specifications of the device are obvious factors. However, there is another factor that is just as important: the decisions made by the team at the right time.
The AI & Physical Devices Product Framework does what traditional methodologies cannot: it organizes these decisions across every stage of product creation, from the initial idea to ongoing improvement.
This approach serves as a guide that helps teams avoid overlooking critical questions and create products in which AI, sensors, and technical components work effectively together.
