Lobster Robotics has developed a cost-effective solution to underwater robotics technology. This start-up has built a small, autonomous robot (called the Lobster Scout) to undertake visual baseline surveys, seabed mapping, and environmental monitoring services. As Lobster Robotics CEO Stephan Rutten says in the interview below: “It’s the quality and scalability of our data that sets us apart.”

Let’s start by asking how Lobster Robotics got started.

We are a group of TU Delft graduates. We had initially been working on a research project to develop a miniaturised camera module for a lunar rover. This concept progressed into Lobster Robotics, a start-up company with a multidisciplinary team looking at the engineering challenges of deep-sea exploration. The deep-sea is actually a comparable environment to space; the conditions are extreme and for the most part unexplored.

Can you tell us the development of the Lobster?

Underwater robotics is very expensive: the high costs are driven by the need for specialised personnel and the day-rate of support vessels. So we started off with a blank sheet of paper to design a new kind of underwater robot, and we soon got quite a lot of commercial attention from companies which were not only looking for a cheaper alternative, but also one that could work in shallower environments. So, the focus for our startup moved away from deep-sea and towards making a tool that can work autonomously in complex coastal environments. This resulted in the Lobster, a 50 kilogram, two-metre long underwater robot equipped with a high-resolution camera.

And what type of tasks can the Lobster do?

The Lobster performs large scale visual surveys of the seabed. The high-resolution photos that it takes are very accurate, with the same lighting and from the same angle. And because single photos are actually not that useful, we create a ‘satellite map’ of the surveyed area. This gives an impression of the whole area which of course is more valuable than individual photos. Because we take so many pictures of the seabed, it becomes much easier to automate the analysis, simply because there is more training data for machine learning approaches. This is where the other big innovation comes in, because currently all the processing is done manually and that takes so much time.

It is autonomous which means it doesn’t have an umbilical. So how long does its batteries last?

Our current Lobster can operate autonomously for three to four hours, depending on the strength of underwater currents. However, next year we are building one that has more endurance: it will be able to work down to 300 metres deep – still in the coastal zone – but can operate in stronger currents.

How important is the fact that the Lobster is autonomous?

Developing an autonomous vehicle helps with the shortage of personnel – this is the biggest reason that we have given the Lobster autonomy. What’s more, sailing survey transects over the seabed is very boring. Companies can use their budgets in a much better way if they use people to do the difficult work, the more challenging specialist tasks such as taking samples, for example.

How does the Lobster know where it is?

This is probably the most challenging question in underwater robotics at the moment: how do you accurately know where you are? You can take the best picture of the seabed but if you don’t know where it is, then it is useless. There are different techniques for this. In shallow waters, we can use the GPS on the robot when it is at the surface while tracking the seafloor with an acoustic sensor. For deeper waters we’re going to need the acoustic underwater network, within which the robot can triangulate its position.

What are your target markets?

Environmental monitoring is our market niche; this means we are looking at clients in the offshore wind, dredging and construction industries as well as governments and knowledge institutions. The Lobster really gives a good picture of the state of the seabed, which vulnerable species are down there, what mitigation plans might need to be made. Our research phase has given us the knowledge and the experience to start working with industrial partners. In addition to our discussions with our PortXL partners, we are also talking to offshore wind developers in the United States to carry out habitat surveys. For example, we’re in the middle of defining a pilot project with Avangrid.

What does the PortXL programme mean to Lobster Robotics?

It is quite hard for a start-up to get the time and attention from the big companies. This is why PortXL is so important for us; to make this time available with potential clients and help us get directly to the right people! We would recommend this programme to any start-up setting up in the marine sector.