An Artificial Intelligence Revolution
Fire Protection at Bosch Building Technologies
GIT SECURITY: Mr. Hirsch, let us start with a brief trip down memory lane: before joining Bosch in 2012, you worked as a management consultant. What exactly were your responsibilities and what experiences did you gain that proved beneficial at Bosch?
Michael Hirsch: I would say it provided me with a strategic and analytical approach as well as international experience across different industries that I can bring to my current role at Bosch and for the benefit of our customers. It enables me to look at things from different angles.
You were most recently responsible for key account management at Bosch ...
Michael Hirsch: That is correct. I was responsible for large and strategic accounts in our system integrator business. I carried responsibility in a holistic sense: from planning and sales to technology and service. Fire alarm technology was of course also an essential part of this role, and I was able to gain a thorough understanding of our clients’ needs. My new role expands the focus to include responsibility for product and solution development in this division.
Could you describe the positioning of fire alarm technology in the Bosch Building Technologies security and safety portfolio? What is its role alongside video security, intrusion detection systems, access control and voice evacuation systems?
Michael Hirsch: Fire detection technology is a very important part of the overall portfolio of Bosch Building Technologies. Our goal is to protect human lives and company assets through the use of our technology. This is also a core value of the Bosch Group with its guiding principle “Invented for life”. But we also benefit from significant interfaces and touchpoints within our division. On the one hand, we work very closely with our own system integrator business; on the other hand, we can internally draw on very broad and in-depth domain know-how.
How do you position your unit with regards to your international competitors as far as fire detection and fire alarm solutions are concerned? You just mentioned domain knowledge. Could you explain how the Fire business unit translates this know-how into customer benefits?
Michael Hirsch: The critical factor is that as well as very innovative customized solutions, we also offer cross-domain solutions. In addition to fire alarm systems, we also have the domains of communication systems, i.e. voice alarm and evacuation, or intrusion detection technology, access control, and video security systems. This allows for a completely different depth of customer solutions compared to solely offering expertise in one area. In the event of a fire, we can quickly and safely direct the affected persons out of the building. In addition, we set benchmarks in our industry, for example with AIoT-based video fire detection.
You just dropped the keyword ‘AIoT’. At Bosch, you use this term to express the merging of two megatrends, namely Artificial Intelligence (AI) and the Internet of Things (IoT). Both are currently among the major growth markets worldwide and, according to Bosch CEO Volkmar Denner, they form a central part of your company’s corporate strategy. Could you please go into a little more detail on what this is all about from the Bosch perspective?
Michael Hirsch: Artificial intelligence, or AI for short, is a key technology for Bosch. Volkmar Denner has made a strong point that from 2025, all Bosch products will have AI, or be developed or manufactured with it. To achieve this goal, Bosch has set up its own competence center, the Bosch Center for Artificial Intelligence, or BCAI for short. The relevance of AIoT is also steadily increasing in fire protection, especially in fire detection. We have seen a clear trend in that segment toward video-based fire detection for years as a supplement to standard fire detection systems. Thanks to the use of AI, we are able to detect more quickly whether a fire is developing or not. This technology works very reliably and avoids false alarms. The algorithms are constantly being further improved for ever greater accuracy and speed. A second field revolves around remote services. Today we can recognize the contamination level of a detector more quickly and we can react accordingly. We will increasingly make use of AI in this field as well, and get better insights for planning maintenance measures and ensuring system stability.
Let us take a closer look at video-based fire detection – the keyword being Aviotec?
Michael Hirsch: This is a very important use case for AIoT. The foundation was initially created here with knowledge-based models: flames and smoke can be characterized, so the corresponding descriptive features formed patterns and led to an initial model. Machine learning was added in the second step. Self-learning systems improve performance, especially the important robustness against false alarm scenarios. Deep learning holds further potential, but since AI also has limits, it is important for us to understand what results it arrives at. After all, human lives and values are at stake. That is why AI currently represents a complementary technology to normative fire alarm systems.
For which solutions do you expect to see particularly strong growth?
Michael Hirsch: Basically, we market our systems through our own installer business, but also through system integrators and other partners. We see very good growth opportunities here, among other things through the new Avenar product generation, as well as remote services and other innovations. The field of video-based fire detection will certainly be very important, and customers will also be able to very clearly see the benefits that go beyond normative fire detection technology. This has already been recognized by the jury of the GIT SECURITY AWARD 2020, where we won first place in the fire protection category with Aviotec. Fast and reliable detection directly at the source of the fire is simply very important for users worldwide. Take one practical example among many, that of a paper processing plant: the ceilings are tall, and it takes a long time for smoke to rise up to reach conventional fire detectors – long enough to turn a small glowing ember into a fully-fledged fire. With our video-based fire detection on the other hand, you get instant alerts when a fire source is detected. That is a significant added value. The same applies to all types of warehouses, especially those that are highly automated and with few employees. Data centers or even tunnels are further examples.
Remote services via an Internet connection, which you already mentioned, are also important here: fire alarm systems can be maintained completely remotely, right?
Michael Hirsch: That is correct. Our remote services bring significant benefits for our customers. Decentralized access allows us to respond quickly and efficiently. This includes remote access to our systems for programming and maintenance, live monitoring of the accruing sensor data, and forwarding of alarms and error messages to smart devices.
You recently launched a special app for that as well ...
Michael Hirsch: Up until now we have focused mainly on enhancing the efficiency and speed of our system integrators. With the new app we are significantly expanding remote services by also bringing them to the end customer. The operator of a building can get access in real time, which represents a big added value for the customer.
Mr. Hirsch, another key topic on a meta-level within the industry of security and building technologies is the connectivity of different components within a security system. For fire protection technology, for example, this includes integration with alarms, emergency lighting, and smoke extraction. What do you have in store at Bosch in that respect?
Michael Hirsch: In this case, we benefit greatly from the fact that Bosch Building Technologies combines various domains under one roof. Our fire alarm systems, in addition to being linked via simple interfaces, can be connected to our powerful voice alarm systems via an IP connection. The system meets the highest standards in terms of quality, security, and functionality, and also offers outstanding flexibility as well as scalability.
Let us talk about some more of the latest innovations in Bosch’s fire protection portfolio. For example the Avenar
series of signaling devices?
Michael Hirsch: At the beginning of the year, we launched the Avenar all-in-one 4000, a combined optical/acoustic signaling device that complies with EN54-23 and EN54-3. It combines optical and sound alarms, allows up to 125 devices on one ring, and is quick and easy to install. Its modularity and flexibility provide a substantial advantage. At public buildings and hotels in particular, this product will be of great benefit.
Are there any other exciting innovations in the pipeline at Bosch?
Michael Hirsch: One milestone is the expansion of video-based fire detection applications to areas outside of buildings. This is an absolute novelty in the industry. We had received a lot of customer feedback indicating an existing need to monitor outdoor storage areas, for example. Due to the varying weather conditions in outdoor areas, this is not a simple task. You have to be able to reliably deal with a wide range of environmental conditions such as fog, snow and wind, which of course do not exist in closed buildings. As I said earlier, the use of AI is helping us to develop the solution further. That is just one example of the many exciting innovations we have in store for the near future.
Aviotec IP starlight 8000
In certain environments, additional protection besides smoke detectors makes sense. For example, if a fire is detected in a room with high ceilings before the smoke reaches the smoke detector mounted high up, it is possible to react more quickly and thus gain valuable time. It is equally important to correctly distinguish fires and smoke from other faults that can trigger false alarms and have financial consequences.
Aviotec IP starlight 8000 makes it possible, for example, in applications in industry, transport and warehousing or even energy supply, to keep the detection time particularly short and minimise false alarms. Thanks to faster detection and increased situational awareness, rescue measures are accelerated, lives are protected and damage is reduced.
The Bosch video-based fire detection system uses a special physical fire detection model that reliably distinguishes between malfunctions and real fires. Thanks to Bosch‘s camera technology and intelligent software algorithms, the video footage is analysed directly in the camera. An additional processing unit is not required, which reduces installation effort and costs.