" is expected to make its maiden voyage in 2017, the product of a $27 million collaboration between the Massachusetts Institute of Technology (MIT) and the Amsterdam Institute for Advanced Metropolitan Solutions (AMS).
The cutting-edge crafts will offer a range of urban solutions from public transport to pollution sensing and water-based venues.
If trials are successful, a fleet of Roboats will soon be operating in the Dutch capital, and around the world.
The autonomous boats will provide an alternative, eco-friendly transport option.
"In Amsterdam, the existing infrastructure of roads and bridges is extremely busy," says Dr. Stephan van Dijk, research program manager at AMS
. "We will try to shift the transport of goods and people to the waterways to get people out of their cars and reduce traffic in the city."
The boats are designed to be light and easy to maneuver, says van Dijk, so that they can quickly and easily reach all points of Amsterdam's extensive canal network, which made the city the obvious location for the project.
The design also allows boats to be connected together to form floating bridges and platforms, that could be used for purposes from disaster relief to open air concerts.
Roboats will be fitted with a battery of sensors to deliver sophisticated data on water and air quality, as well as information on pollution -- including the numerous bicycles that fall in the city's canals each year.
"Air quality measures usually come from a few places in a city that do not tell you much about human exposure," says Professor Carlo Ratti, Director of the MIT Senseable City Lab
. "The next level is to monitor in a ubiquitous way with a lot of fine-grained measurement...which we can do with this platform."
The team hope this data could inform policy to improve air and water quality in Amsterdam, and eventually in countries such as India where dirty water is a deadly threat
Under the hood
The vessels make use of techniques developed for autonomous cars, although the demands are different.
"The technology is like in self-driving cars with sensors scanning the environment and building a three-dimensional model to navigate," says Ratti. "The key difference is our sensors need to look not only at the surface but also what is happening under the water...the algorithms and artificial intelligence are different but the principles and overall architecture are quite similar."
Perfecting these systems will be a laborious task, which is why the project will last five years despite the first prototype hitting the water in 2017.
"We need to have enough learning," says Ratti. "Just like with self-driving cars we want to accumulate many miles to make sure the system is safe, and to test it for an extensive period of time."
Navigation on the water will create new challenges, according to Sally Applin, a Silicon Valley-based member of the Internet of Things council and a doctoral candidate at the University of Kent studying humans and automation.
"The most critical factor here from an engineering standpoint... is the required sociability between Roboats so that they will know how to cooperate with each other and with non-robotic boats sharing the canals," says Applin. "How will Roboats be able to negotiate with each other to navigate a medium that has less rigid 'lanes' and rules and in the proximity of other boats?"
The field of self-driving boats is likely to expand in the coming years, with several companies developing new designs.
Rolls Royce has announced plans for an autonomous cargo vessel
, although this has faced opposition
from the shipping industry over safety and piracy concerns.
But the relatively calm waters of Amsterdam could offer the ideal proving ground for autonomous vessels, says Ajit Jaokar, who teaches "Data Science for the Internet of Things
" at the University of Oxford.
"I do not see the same (safety) issues here," he says. "As with all AI applications, Roboat will get better as it navigates more waterways. Once a model is developed, it could train other boats."
Professor Ratti says the project has received interest from cities around the world, eager to apply the technology.
He also believes the system could be used with many other types of vessels to reduce fatal accidents at sea.
If the Amsterdam project is successful, the history of autonomous vehicles could receive a significant update.