If you believe IoT, machine learning, and artificial intelligence are the driving technologies shaping the future of video surveillance, then a recent event held by Milestone Systems was the place to be. Billed as ‘Creating an Intelligent World Hackathon’, the gathering was held near Copenhagen on the eastern coast of Denmark.
With the goal being to push the boundaries of video technology, Milestone brought together young technology and solution developers for a weekend of intense innovation and competition. Students, professionals, and industry experts searched for breakthrough ideas to spur new ways of solving problems with video technology.
Combining use-case presentations, tech talks, teamwork, and of course actual coding, the Hackathon offered a unique opportunity for developers to work with some of the biggest players in logistics and retail to create leading-edge solutions to business challenges — and a chance to win 25,000 Danish kroner, in cash.
An Intense Weekend
With 28 young people, six mentors and five judges — all equipped with little more than sleeping bags and computers — the stage was set. At 9:00 a.m. Saturday morning, these software, business, and UX developers spent the next 24 hours exploring real-world problems and envisioning new concepts for future video technology solutions.
Milestone Systems CTO Bjørn Skou Eilertsen welcomed the group and challenged everyone to create a more intelligent world by using video technology, IoT and machine learning. Jørgen Schur, Nordic Leader, IBM Watson Customer Engagement, spoke about current problems and future technology in retail, followed by Berian James, Maersk’s Head of Data Science and AI, who discussed the impact of IoT in the global logistics sector.
Seven teams were assembled, and the competition was underway. Quickly, meeting rooms and hallways were abuzz with everyone intensely discussing, white-boarding, working through potential solution scenarios and writing code, as mentors from Maersk, Easysize, Nordeus, IBM and Milestone shared advice and guidance with the participants. There was not much sleep that night.
The Winning Solution
When the sun rose on Sunday morning, participants were ready to pitch their ideas for a panel of judges consisting of Milestone Systems, Maersk, IBM, GE Digital and Design Matters. Each team had five minutes to present their idea, and by the afternoon, all seven teams were waiting excitedly to hear the verdict.
The winning team, named Trackport, had developed an application that, through predictive analysis technology, could reduce the pilfering and theft of goods from trucks and vans that were waiting for long periods of time to load goods onto container ships in industrial ports around the world. The time and money lost to loading-dock theft is quite significant. Port Elizabeth in South Africa, for example, loses an estimated $15,000 USD per day.
The Trackport approach to the problem was straightforward: User data from individual vehicles is combined with video data from the port area and analyzed, after which the truck drivers can see how long the dock loading queues are and decide on a time of day that’s best to drive to the port for offloading. By reducing the downtime of waiting in a queue, the trucks have minimal exposure to pilfering. The Trackport team estimated that Port Elizabeth could achieve 44 percent cost savings.
There were many good bids so it was hard to choose, but Trackport’s idea used machine learning to extract relevant information from video data. Trackport offers a serious solution to a widespread problem in the transport and logistics sector. Their business case was solid, and we believe in the possibilities. – Bjørn Skou Eilertsen, CTO, Milestone Systems
Congratulations to the Trackport team:
- Mathilde Fynbo, s student of Digital Innovation and Management, ITU
- Amelia Nielsen, a Digital Innovation and Management student from ITU
- Krisjanis Vaiders, a developer from the startup company, IndieFrame
- Bertalan Kovacs, a student of Autonomous Systems, Aalborg University
- Andreas Klingenbrunn, a student of Data Science at ITU
For more details see the video and bios:
See also information on the global Milestone Kickstarter Contest: