According to Bosch, anyone who connects the Internet of Things with artificial intelligence starts a cycle in which products and services create more and more value for everyone involved. How does that work?

What do you get when you count AI in IoT? A nice anagram, you will find at Bosch, but also a value-added motor that reinforces itself. The technology specialist calls the merger “AIoT” and sees it as a key component of its future strategy. “With AIoT, we naturally want to make our devices more intelligent, but we also want to better include the perspective of our users in our product development,” says Michael Fausten, head of the AI ​​and Systems division at Bosch.

AIoT cycle

He speaks of a value-added cycle that results from the combination of AI and IoT. According to Fausten, the concept is simple: “Connected products generate data that we can analyze with AI and other methods based on machine learning. The resulting knowledge allows us to update our software and products and improve functionality.”

Bosch sees a cycle consisting of four phases:

• added value: Bosch uses insights from data in its R&D process to improve applications or devices.
• products and consumers: The connected products are used by customers and thus generate (anonymous) data.
• data processing: The data ends up in the cloud, where it is stored in a structured manner. Bosch ensures that the necessary privacy is integrated here.
• AI Algorithms: AI looks at the structured data generated by connected IoT devices to gain new insights and make products better and safer.

Bosch has no shortage of telemetry to make this model a reality. The manufacturer is responsible for a large number of devices. Don’t just think of washing machines or refrigerators for consumers, but also machines in factories, smart cameras and sensors for (semi-autonomous) cars. When these devices are connected, Bosch can work with the sensor data. Although, of course, this raises questions about security and privacy.

Analyze for sure

Fausten notices that too. “Products in the B2C context are networked even before they reach the customer. The data they collect and send back is therefore completely anonymous. Reliable handling of the data is a matter of course for us.” Bosch then uses the collected data to adapt and improve products, also during their lifetime. “With AIoT, we will develop products together with our customers in the future. This is a radically different approach than before.”

In a B2B environment, Bosch deploys AIoT together with the customer via its AI Analytics platform. This scalable platform analyzes billions of data points to find solutions to complex manufacturing problems in near real time, says Fausten. “By combining AI with telemetry from machines and the production process, Bosch can detect anomalies and possible errors in good time. In this way, experts can intervene before production actually comes to a standstill.”

Smart and therefore efficient factory

Bosch itself relies on this technology at its plant in Dresden. All data from machines, sensors and products are stored there in a central database. Every second, the factory generates data equivalent to five hundred pages of text. AI looks at this data and ensures that processes are optimized in real time without interrupting production. The AIoT approach also ensures that Bosch can quickly detect and fix anomalies.

Through machine learning on IoT data, the company gains insights that optimize the production process. According to Fausten, this leads to better products, more efficient and sustainable factories and ultimately happier customers. If Bosch customers themselves rely on the AIoT cycle, they can reap the same benefits, he says.

Essentially, AIoT creates a value cycle where products improve themselves by analyzing their telemetry. Preventive maintenance is a simple application, the cycle can also make equipment and machines more reliable over time and common faults go away. This self-reinforcing cycle, where AI and IoT data leads to better and smarter products, is a key part of Bosch’s current strategy.

Smart camera gets smarter

Bosch already has many examples of the AIoT cycle outside of its own factories. In the field of smart cars, for example, Bosch offers cameras that see and interpret the environment. The introduction of AI into this telemetry ensures that the camera not only sees the behavior of road users, but can also predict it. For example, a roadside pedestrian approaching a crosswalk will most likely want to cross.

The AIoT cycle here is as follows: cameras see the environment, anonymous data then goes to Bosch’s cloud environment, AI analyzes the data to improve predictive behavior, this knowledge returns to the cameras and history repeats itself . Fausten emphasizes that the added value from data analysis not only benefits the manufacturer, but certainly also the customer.

We see the same principle being used in on-board computers for e-bikes. You can predict a cyclist’s arrival time based on the route, but also other factors such as the type of bike and the condition of the road. By sending anonymized data back to the cloud and then unlocking AI on top of the IoT data, the on-board computer becomes smarter.

Fausten believes AIoT is the key to future value creation. “Adoption of AI completes the cycle,” he reiterates. “Product development and improvement can go so much faster. We can be disruptive.”