Autonomous vessels are a complex combination of IoT, ICT, AI and state of the art image recognition technology. The latest demonstration has given the green light to move forward with the most difficult obstacle which is developing AI image capture that can detect other ships and ultimately allows for safe navigation. Mastering the technology to dock a vessel, which is considered the most difficult manoeuvre, depends on perfecting AI with image capture technology.
Satoru Kuwahara who is DFFAS (Designing the future of fully autonomous ship) consortium’s project director says short range navigation planning will be controlled by the autonomous vessel rather than the team at the control centre. This means having confidence in the integrated technology onboard for safe navigation. “There is still room for improving algorithms so they match the skill level of humans in terms of short-term manoeuvring.”
Embracing fully autonomous vessels also leaves them vulnerable to electronic hijacking and pirates. Mr Kuwahara stresses the biggest priority is protecting the lives of crew members which is already guaranteed by going digital. Preventing cyber attacks play a significant role compared to conventional ships. Communication with ships will be backed up by satellites and IT security software inbuilt to the systems on land. But Mr Unno likens the challenge to climbing a mountain in which the project has only passed the halfway mark.
Japan has the world’s second largest shipping industry and third largest shipbuilding industry. But it risks being overtaken by other rising economies. Mr Unno believes Japan has a global headstart developing autonomous shipbuilding technology. But he concedes that every step of the way could be plagued by setbacks.
“We still need to do more and more simulations, and we also need to put the simulations to practical use on actual shipping routes. We need more data which will increase the number of opportunities for AI to learn from this data.”