Top Picks

Here are our current top recommended tools, gear, and kits for scientists working with machine learning, data collection, and field research.

Laptop for ML/DL/Advanced Computing

Alienware M18 R2
This isn’t our TOP, top pick…but for the money, this machine rips. It should be capable of handling most local training/development tasks, is capable of handling heavy processing tasks with the i9, and has enough memory to tackle larger data sets. You wouldn’t be sad having this as your local powerhouse.
Link

Field Drone

Autel EVO II Pro RTK V3
This sUAS checks a lot of boxes for us. Its RTK-equipped out of the box, RID compliant, supports native mission planning, has 360-degree obstacle avoidance, and has an acceptable flight time. When you are investing $3000 in a drone, this level of features is a lot of value. You can certainly spend the same or more and not get features like RTK.
Link

Raspberry Pi Kits and Accessories

*It’s hard to nail down our favorites in this vast space. However, we recommend that you as ecological and environmental researchers start experimenting with Raspberry Pis, Arduinos, and any other micro-computers you can get your hands on. The potential these have to streamline workflows in our field is immense. From autonomous data collection to mesh networking of sensors…Raspberry Pis will transform how you collect and interact with data.

Raspberry Pi/Arduino Outdoor enclosure
Now, this isn’t waterproof, but it’s close. If you’re truly looking for waterproof, we recommend a custom case build from a certified waterproof enclosure. Link

Raspberry Pi Sense HAT An OEM Raspberry Pi add-on board that is almost an automatic for any sensor unit we build. It facilitates the monitoring of pressure, humidity, and temperature (among others). Interestingly enough, it also has a variety of position and movement sensors. One way that we have found that to be useful is setting up an alert on changes in the accelerometer to monitor disturbances to the sensor unit remotely. It’s a plug-and-play board that is a great addition to any Pi project. Link

CanaKit Raspberry Pi 5 Starter Kit MAX CanaKit has been around for a while and offers some great off the shelf solutions. This kit, with the new RPi 5, 16GB of RAM, and an included 256GB MicroSD card…is a giant in a tiny package. It is certainly capable of running a lite ML/DL model (wrapped with TF Lite or PyTorch Mobile) on board. This unit could record and store a significant amount of tabular data and supports networking add-ons (4G HATS, which also frequently offer GNSS positioning). Couple all that with integrated cooling and this unit offers most of what you could ask for. Link

**Home/Lab Gear We call it a Home/Lab. We like to say the line between home and lab is blurred, especially when you have remote access. This could be its own list, but we just want to get you thinking. Start smaller, open up your horizons, then you can turbo charge your research life with a powerhouse. Refurbished servers are a great option to get into your own data center, so are old PCs. It just depends on your workload. Think about a NAS (Network Attached Storage, see below) and then go from there. We like a NAS unit that offers virtualization (think Docker) and RAID support. Look for our DIY homelab build for environmental and ecological researchers article soon. Synology DS1821 NAS