Doctor of Energetics Heigo Mõlder, a member of the development team, says that the long-term goal is to build a modular robotic platform that allows for various activities without human intervention. The specific application is not yet firmly determined.
The first launch of the prototype this summer in the waters of Männiku quarry was preceded by a couple of years of preparatory work to find funding and partners to ensure that the invention thought up in laboratory would be technically implemented and had a practical application.
However, the idea for the robotic vessel originated from the wheeled robotic platform that can be controlled via a smart device, which was developed in 2016 by Tanel Jalakas, a senior research scientist of TalTech as well as a member of the development team. As similar solutions by others were already essentially finished at the time, and the debut of parcel robots on the streets of Tallinn was only a few months away, it was decided that there would be no point in competing in this area.
Thus came the idea of putting the existing software solution on water. While it seems easier to build a sea-navigating robot than a land-navigating one, there are actually dozens of things on the sea to consider when developing the “brain” of a robotic vessel – such as recognising objects in fog and bright sunshine, seeing underwater objects, complying with maritime traffic laws, etc.
In 2016, members of the TalTech development team brought the concept of a robotic vessel to an event called TehnoHack. It is a two-day hackathon where young inventors come up with ideas, a panel of industry experts evaluates them and selects the best ones to be funded and realised within two days. As a result of TehnoHack, a small robotic boat was built, able to travel from point A to point B according to predetermined coordinates.
Heigo Mõlder then wrote down a longer vision to introduce the idea of a self-steering vessel and the robotic platform in general, made a presentation and delivered it at Robotex robotics festival, the Federation of Estonian Engineering Industry and elsewhere. As a result, he was successful in attracting the attention of engineering company MEC which designs smaller crafts.
The University of Technology has expertise in power electronics, mechatronics and control systems, while MEC has a strong track record in marine mechanical engineering, and the decision to join forces was thus made.
Via the Federation of Estonian Engineering Industry, a sheet metal processing company Hyrles became also interested in the project. As the parent company of Hyrles also produces mass-electronics in Finland, they were looking for a way to develop their own innovative product.
“We liked the idea and combined the three competencies in order to make a robotic vessel platform. The idea was to make it as modular as possible so that the platform could be easily adapted to its purpose at a later stage,” says Heigo Mõlder. “After that, the entire thing was concrete enough and we were able to submit an application to Archimedes for a grant to carry out applied research in smart specialisation. The response was positive and now we have been working since the beginning of 2018 to realise the first goals of the vessel’s prototype creation, which is to achieve the functionality that was initially determined.”
The developers of the robotic vessel also collaborate with the Small Craft Competence Centre of Tallinn University of Technology in Kuressaare, the specialists of which helped to design the vessel’s hull and carry out the first buoyancy tests.
First, an universal base solution
Within the framework of Archimedes’ applied research, a universal solution with as many features as possible and suitable for many activities will be created. Probably not all of the features will be needed later, and the vessel that is actually, so to say, launched in the waters of practical application will be technically simpler than the prototype.
This phase will last until March of next year. The goal is to build a complete, fully electric robotic vessel that can autonomously navigate both in harbour and at sea, avoid obstacles, complete a given mission, is aware of its energy consumption and reserves, is able to return on time and can be manually controlled by radio, if necessary.
“After that, we will focus on specialisation – whether we start transporting rubbish or parcels from the islands to the mainland, or why not both, or the vessel starts mapping the seabed or monitoring the sea, looking for pollutants discharged into the sea, etc.,” Mõlder says.
All the ingenuity that helps to operate the robotic vessel lies in the on-board black box which has been developed by the TalTech development team. Its manager is Indrek Roasto, while the project is led by Mart Enok on the side of Hyrles OÜ.
Heigo Mõlder says that the most difficult part is developing the ingenuity, which in many ways consists of programming and integrating different sensor systems. Of course, the mechanics and the vessel’s power system also need to be dealt with, but these jobs are more technical and can be done via abundant testing.
Automatic navigating systems have been created in the world before, but most of them are at the disposal of a military of some country and their nature is veiled in secrecy. Some solutions can be purchased, but they cost hundreds of thousands of euros, so it makes more sense to design one’s own system according to one’s needs.
The first trial – a sea voyage to Keri Island
As said, truth comes from experimentation. Thus, the sea voyage to Keri Island at the beginning of September became the first real so-called sea trial for the robot vessel with a length of 2.5 m and weighing a little more than 100 kg. During the voyage to Keri and back, members of the Nymo development team collected data and observations on the vessel’s navigational capability and technical resistance in offshore conditions. The vessel was exemplary when navigating to the destination and back, using predetermined waypoints. However, after reaching the end point marked on the north side of Keri, the ship made a sudden turn, throwing an axle bushing out of a bearing, and that took out one of the propellers. It had to be repaired on the island, but it did not result in anything catastrophic.
“In terms of electronics and software, everything went according to plan, there were small hiccups in regard to mechanics, but nothing that could not be developed further,” concludes Heigo Mõlder the voyage. “The robotic vessel did a great job and we in turn received a lot of data to analyse further.”
For example, when the robotic vessel crossed paths with several sailboats sailing to Prangli Island, their routes were later retrieved from the Automatic Identification System (AIS) for identifying ships, and software developers were able to analyse and simulate various situations on how the sailing routes of the sailboats and Nymo might have crossed.
By transmitting information from AIS and Marine Traffic (a global ship tracking application in which all ships over 12 meters in length must register their location) to Nymo’s controller, it can predict the routes of the ships that are on its course as well as their estimated crossing place and time in regard to Nymo’s route. To do this, the vessel needs to take two waypoints from the sailing routes of a ship, find the time it took to navigate them, calculate the speed of the ship and use the data to set its own route so that it does not meet another ship.
“We are halfway through the development of this so-called advanced software. In principle, the robotic vessel should automatically pick up the necessary waypoints of other ships and ideally choose a course according to information from the other ships before the trip has even begun,” Mõlder describes. “But since forecasts are never perfect, the control system must be ready to change course according to the situation encountered while travelling.”
An autonomous robotic vessel must be created
At a higher level, the team wants to enter nautical charts, with information such as depth, larger rocks, buoys and the coastline, to the system so that the vessel can take into account the particularities of the route before travelling. For example, when the robotic vessel begins to travel between the mainland and Aegna Island, it already has all the information which needs to be taken into account on this route. This route can be then locked and the vessel no longer needs, for example, a seabed sensor or other sensors, which allows making it lighter in weight, but also makes the system easier.
Due to the peculiarities of hydrodynamics, the energy consumption of the vessel responds non-linearly to the increase of the vessel’s speed. For example, if the robotic vessel is travelling at a speed of three knots and its energy consumption is X, then as the speed increases to six knots, the energy consumption does not increase twice, but three times. Much will depend on the waves and if the vessel travels very quickly, the hull will collide with the waves, the contact with water will decrease, the water resistance will be reduced and the energy consumption per the voyage’s length will also drop.
All these factors need to be taken into account and the optimum choice has to be found considering the loading of the ship, sailing route and the speed. The most efficient speed on the current vessel is 3 knots or 5-6 km/h, with a maximum of 7 knots (13 km/h). If it were to travel to Aegna Island, which is located 14 km away from Kalasadam, and then a one-way voyage would take 2.5 hours. After all, parcel robots are not driving more quickly on the streets.
The aim is to bring the ship to a level where human intervention is minimal. One option in consideration is the capability of the vessel to start towing a barge bearing a rubbish bin from a harbour as soon as the bin has been filled.
The choice is between electric and hybrid motors
Whatever tasks the robotic vessel will perform in the future, attention must be paid in phase II to carrying capacity. The aim is for the vessel to be able to carry up to two full European pallets. A catamaran-type hull is probably the best solution.
In principle, it is possible to place the aforementioned black box on any recreational craft, harmonise the radio communication, add steering mechanics, etc., and the vessel should become self-steering. Of course, the tonnage and other parameters of the vessel need to be recalculated in the software, but for the programmer it is mainly about changing the setup, because the information from the sensors would already be tested.
There are two options for choosing the power source – whether to keep using battery power or to use a hybrid motor with a diesel-electric generator. There is no difference in regard to the technical structure, but the electric motor would enable to make the ship fully automatic. Someone still has to fuel the tank, but the recharging of the battery can be made wireless, and thus fully automatic, by using inductive coils on both the ship and the quay.
“In addition, the efficiency of electrical systems is higher than that of diesel drives, and that’s why the future of power sources is predicted to be electric,” Mõlder explains. “At the same time, the problem with electricity is that battery technology is lagging behind our needs and it is relatively expensive to implement high-capacity solutions. In this sense, a diesel engine would be several times cheaper.”
In parallel with everything else, scientists must also find funding for phase II of the robotic vessel development project: for product development. Work is currently carried out with the goal that once the prototype of the vessel is ready, Hyrles would take responsibility for its mass-production.
In order to announce their activities and attract interest in those concerned, a survey is conducted among the permanent residents of the small islands through the Association of Estonian Islands to find places where it would make sense to start transporting parcels by sea as a pilot project. When Heigo Mõlder was presenting the project, representatives from 18 islands came to hear the presentation.
“Initially, it would be good if we could test the ship’s functionality, durability, mobile applications, how they work and justify themselves in the pilot project,” says Mõlder, thinking it necessary to give the vessel a practical application. “The ultimate goal is for us to reach a Nordic market where the needs and opportunities are much greater. And in order to prove ourselves there, such references are important.”
Hyrles would be glad to start producing the robotic vessel
Urmo Sisask, Hyrles OÜ, Executive Director:
Our company is interested in new technologies, products and solutions that make various activities easier, more comfortable and safer for people. The goal when joining the project was to be the manufacturer of the final product, the robotic vessel, and implement new necessary technologies and develop skills in our company. We want to create a platform that is as universal as possible and can be later used in many areas.
In regard to mass-production, it is currently difficult to give an assessment as no sales have yet been made, but there are interested parties. There are plenty of applications for such a robotic vessel. Examples include seabed mapping, monitoring of fish stocks, counting birds or, for example, seals, marine pollution monitoring, plus marine rescue, border guard and solutions regarding defence. In the Nordic countries, mail and parcel services to small islands may be worth considering.
The vessel’s development is on the right track
Meelis Mäesalu, Director of company OÜ MEC Insenerilahendused:
Autonomous and unmanned vessels can be the next big step in the shipping and marine industries. As an engineering company, we need to stay abreast of technological developments, and that was one of the main motivations for the development of a robotic vessel. Estonia could be a suitable country to develop and test a robotic vessel, because we have the competence in the respective disciplines and the communication with various organisations and institutions is flexible and fast. In addition to acquiring knowledge, we hope that, in cooperation with our partners, the robotic vessel can become an independent product.
Of course, there is room for unpredictability. Such a robotic vessel has many aspects that are not encountered in classic shipbuilding. Energy sources, innovative control systems and small-scale propulsion systems have resulted in many surprises.
The ship has made several voyages on its own, following a predetermined route, so the development is on the right track. The next big step is to ensure that the vessel is able to understand the area, behave adequately, taking into account the circumstances and rules of the sea, and perform the function assigned to it. This means that in addition to complying with the commands, the vessel must be able to make its own decisions.
A promising initiative for Naissaar Island
Olar Gofman, member of the board of the Association of Estonian Islands from Naissaar:
After the development of land-navigating robots, this is a logical continuation of things and developments. It could be significantly beneficial outside the navigation season when most regular boats are lifted out of the sea and there is little movement between the island and the mainland.
There are about a dozen people living on Naissaar Island all year round. Providing them with food, bringing mail and carrying out other similar activities, as well as taking the rubbish on the voyage back, for example, could certainly be topics for further discussion. But there are even more such areas. Much depends, of course, on the carrying capacity of the vessel and its seaworthiness.
Although the topic has not yet been discussed in the community board, I personally believe we would be interested in participating in this project.
Robotic vessel Nymo
- The vessel is being developed by companies OÜ Hyrles and OÜ MEC Insenerilahendused in cooperation with the Department of Electrical Power Engineering and Mechatronics of Tallinn University of Technology.
- The goal is to create an autonomous self-navigating robotic vessel.
- Technical specifications of robotic vessel Nymo:
- length 2.5 m
- width 1.1 m
- carrying capacity 100 kg
- displacement tonnage 250 kg
- power source – two electric motors 1.5 + 1.5 kW
- draught 0.36 m
- maximum speed 7 knots (13 km/h)
- operating speed 3 knots (5.6 km/h)
- travel time at operating speed 24 h
- travel range up to 100 km
- towing capacity 500 kg
Source: Tallinn University of Technology