Aardvark Sensing BV

About Gamma Ray sensors An important constant attribute of soils is the texture, the ratio between different grain sizes (Clay, Silt, Sand), This ratio will tell us a lot about carbon holding capacity, water holding capacity and availability of nutrients and chemicals to the plant. One great way to classify soils without digging and taking it to the lab is by using a gamma-ray sensor. Clay, for example, tends to collect and retain more of the naturally occurring Potassium-40 radioactive isotope. This means that a gamma ray sensor can detect the decay of that isotope. Gamma-ray sensors are used either on vehicles or on drones, and of course are a good candidate for mounting on a robot. Due to the slow driving pace of the Aardvark Robot (Aardvark Sensing BV), we can afford to use a small (and less expensive) sensor while achieving good sensitivity and great resolution. In the map below, you can see just how much variance there is in soil texture even within a single field. This stresses the importance of soil sensing in the context of agriculture. This is also an opportunity for a shout-out to our colleagues at Medusa Explorations BV, who make and use those sensors. ESA Business Incubation Centre Noordwijk Image source: https://lnkd.in/e63ZiSKR https://lnkd.in/enibReJ9

Last week at the Smart City expo in Barcelona, I had the pleasure to meet with Alf Olsen . I found that Alf and I share a lot in our vision of robotics. Alf is the founder of O-Robotics, a Danish startup that makes autonomous robots for various maintenance tasks in urban areas. Best of luck Alf! Aardvark Sensing BV

Some more torture for Aardvark SN2 It has survived the wet salty sand on the Katwijk beach, so we went on trying to kill it some other way. SN2 is most definitely not IP67 rated, but when we saw this big puddle outside the ESA Business Incubation Centre Noordwijk we had to drive through it (what would spaceX do, right?). Well, it turned out the puddle was deeper than expected, which was very funny. SN2 was almost completely submerged (note to self : SN3 needs to float), but somehow it maintained radio contact and Ton van Dijk was able to drive it out safely. So great test, it got out on its own power, but that doesn't mean that it didn't ingest water, we'll need to take it apart to find out

A simple test for plant biodiversity in a field As part of our efforts to generate scoring modules for environmental practices, one of the simplest (but valuable) measures for a soil health is biodiversity ☘ What is biodiversity? Simply put - the number of plant species in an area, this indicates the health of the soil, and also can sometimes tell us if herbicides are being used 📷 How do you measure it? We start with a simple visible light camera, pointed downwards from the robot's payload bay 🍃 How do you quantify it? Use machine learning to identify plant species, Use shape and color (currently we only use visible color, but later will integrate with IR bands) 📊 What are the statistics? Provide a histogram of percentage soil surface coverage, from bare soil to various grasses and weeds to crops

SN2 almost ready for navigation testing, exciting!

We identify 4 different roving patterns for earth rovers: 1. Repetitive area survey: A rover is scanning a given grid repetitively every day or every few hours, to produce a 2D time series. This is useful mostly for agricultural research and ground truth collection for satellite AI 2. Single area survey: A rover is brought in to perform a one-time 2D scan of a grid - this is useful for soil type mapping and applications like underground mapping with GPR, and for landmine and EXO detection 3. Linear monitoring: A rover is periodically scanning a long test line, usually every day. A rover can cover several kilometers per day. This is useful for applications like levee integrity monitoring and early detection of piping 4. Exploratory: Much like Mars rovers, in this case a rover carries its own solar panels and uses satellite communications. The rover is completely independent can carries out a month-long journey, usually in uninhabited domains. This is useful for applications as gas and oil and mineral mapping, carrying sensors like Gravimeters.