Singapore Polytechnic collaborates with Bosch to give students hands-on AIoT experience
July 2021
The partnership seeks to demonstrate Bosch Intelligent Video Analytics and machine learning with 5G network connectivity to create real-world use cases that improve safety, business operations, and customer experience while preparing students with relevant technical skills for the future.
Across Singapore Polytechnic’s campus – inside an airport hangar, overlooking a car park, at the canteen – dozens of Bosch intelligent video cameras are being deployed to demonstrate the power of built-in video analytics, a form of artificial intelligence (AI). Students pursuing a degree at Singapore Poly in Electrical & Electronic Engineering (EEE) receive hands-on training and develop AI based computer vision and machine learning skills. The campus serves as a proving ground for real-world AIoT – the combination of AI with the Internet of Things (IoT) – applications bringing together students and industry to solve today’s challenges and anticipate future developments.
“We see a trend towards the use of video cameras as a generic sensor like our eyes,” said Tan Chee Seng, Director of the 5G and Artificial Intelligence of Things (5G & AIoT) Nerve Centre at the School of EEE. “With artificial intelligence, the camera will be able to capture videos and make use of software algorithms to decode what it is seeing. This will open up many new application possibilities in areas such as facial recognition, colour detection, the tracking of object movement, behaviour monitoring, and a lot more.”
To ensure that it maintains its capabilities and know-how at the cutting edge of video analytics, the school embarked on a collaboration with Bosch. The 5G & AIoT Centre is working with Bosch to demonstrate use cases such as detecting fire or smoke in an airport hangar, enforcing parking restrictions in a car park, taking attendance during an examination, and counting people within the canteen to avoid overcrowding and enable safe social distancing.
“We want to be at the forefront of the technology and its applications, and to do that we have to learn from the industry leaders who are also close to the end customers. This ensures that our curriculum is up to date and that our students are equipped with industry-relevant skills,” said Chee Seng.
In 2019, the school launched its first Train-the-Trainers programme. Fifteen teaching staff members visited the Bosch Experience Centre and attended a video analytics workshop. Topics covered camera fundamentals, data analytics processes and algorithms, and industry use cases. The programme also included a hands-on session with Bosch’s Camera Trainer, based on machine learning technology.
Built-in AI capabilities enable Bosch cameras to understand what they see and add sense and structure to captured video data with metadata. This process is an important first step in converting video data into actionable insights and building predictive solutions, helping users anticipate unforeseen events and prevent them from happening. According to Sudhir Tiku, Vice President, Asia Pacific, Bosch Video Systems & Solutions, one feature of the Bosch camera is that it is very precise and granular in metadata generation.
“The camera by default under-stands the size of an object, whether it is a person, or a car, or a bike, or a truck. It also recognises pixel changes in a scene very well, for example, how the pixels change when a person moves around. If people are walking around in a camera scene, you can instantly know how many people there are in the environment, whether someone is loitering, or whether the person deliberately left an object behind.”
The Centre worked with Bosch to co-develop video analytics training materials used in short courses for adult learners and an elective introduced to full-time students in October 2020.
There are plans to embed video analytics into the core IoT curriculum as well. “Video cameras are going to be widely deployed. If you look at digitalisation roadmaps, a lot of use cases require the camera,” said Chee Seng. “We see video analytics as an important area of knowledge for the future generation.”
Video cameras with built-in AI, like Intelligent Video Analytics from Bosch, often span large sites or city areas and require cellular connectivity. For outdoor video surveillance applications like these, 5G can offer near-elimination of latency, boosting camera responsiveness and real-time monitoring. The ability to define network ‘slices’ means that 5G can offer dedicated portions of network capacity for specific applications, eliminating the need to share capacity with multiple us-ers.
Faster speeds and greater capacity brought by 5G networks will enhance the role of video analytics. Working with Bosch, the 5G & AIoT Centre aims to create additional use cases showing how 5G can leverage AIoT solutions to address future technological challenges. For example, video-enabled autonomous surveillance vehicles and flying drones for building facade inspection are equipped to stream live video remotely and wirelessly as well as extract and send video metadata real-time with low latency.
Singapore Poly’s collaboration with Bosch offers a unique opportunity for staff and students to explore emerging technologies, solve real-world problems, and turn ideas into action.
Trends & Technologies
If you are a user, buyer or planner of safety, security or communication solutions, then this is your new information hub. We created this trends and technologies to generate insights, to provide information about useful trends and to support you in creating ideas for better and more efficient solutions.