Klaus Grouleff, Head of Software Development, attends the summit together with Maulin Gajjar and Sandeep Akhouri from Software AG as well as Jabu Sibanyoni from Amazon Web Services (AWS).
Edge computing brings computing closer to its data source, which makes it very relevant when very low-latency computing or data processing is required. Unlike cloud computing, edge doesn’t require a data center or remote server network. Some edge applications are dependent on cloud for e.g. storing data and machine learning, but will for some processes process data close to where it is generated. The benefit of cloud computing is, in essence, that it is less dependent on a strong connection since it requires less data to be processed. In this session, it is discussed how AWS’ edge computing solutions can be utilized in moving data as close to the endpoint as possible.
The summit also presented a chance for the company reps to present use cases on edge computing within their fields. In the use case, Klaus presents how Ubiqisense’s social distancing function was put to use in CITY2, a Danish shopping mall, and the challenges and solutions resulting from it. He also discusses in detail the architecture behind the solution and the outcomes of it.
The pandemic posed a lot of implications for Ubiqisense since virtually no-one was in the office, with all employees except those with positions critical for the country’s continuation sent home. This was a big issue for Ubiqisense since we need to have people in spaces to have a data foundation. However, we came up with the idea of adding a social distancing feature to our product offering. This was tested out over two weeks in CITY2, a shopping mall in Denmark, even making Danish TV.
Before covid-19, Ubiqisense’s solutions mainly served customers in offices and educational institutions, and both were empty for large parts of the first year of the pandemic:
“When covid-19 came and no-one was in the office or educational institution, we asked ourselves if we could deploy our algorithms on existing IP cameras to: 1. help monitor safe distance and crowds to help open society in a controlled and safe way; and 2. increase our product portfolio further than buildings into smart cities, large and open spaces”, says Klaus Grouleff.
With the shopping mall use case, Ubiqisense was challenged by retailers not having sufficient bandwidth to send the video feed to the cloud, which made collecting data without latency very difficult. It was, therefore, necessary to do it on site. Apart from that, the solution needed to be easily scalable, if, for example, the mall installed new IP cameras. It was found that the performance of the cameras depended on the model used, and the models, therefore, needed to be re-trained to be scalable.
To learn more about how Ubiqisense used edge computing to monitor social distancing in CITY2, watch the video here: https://cutt.ly/yQzKEXw
