UKHASnet Hack Weekend

Individual node descriptions to follow in future posts, pictures from Phil Crump and Dave Brooke

After much discussion and very little planning a group of 10 people meet up in Southampton for a weekend of hacking UKHASnet nodes. The plan was to spend Saturday building, programming and setting up UKHASnet nodes and then on the Sunday to head out and field test them. Before the weekend it was agreed to focus on creating a flotilla of UKHASnet buoys.

Saturday

ukhasnet buoy testing

On Saturday we all met up at the University of Southampton’s Engineering Department and set about working on build various UKHASnet nodes with a general approach of a sea buoy. Example nodes include AI2 (GSM/UKHASnet Gateway), QUAD0 (Quadcopter mounted node) and PFLOT0 (APRS/UKHASnet Gateway). Most of the morning and afternoon was hardware hacking and then in the evening we move onto building the ‘hulls’. We investigated a number of methods of waterproofing the nodes including jam jars, ‘tupperware’ food containers and water bottles. What was particularly difficult for us to get our heads around was making nodes heavy rather then the traditional push towards low mass for high altitude balloons.
By the end of the evening we have successfully tested (in the sink) 4 nodes which we hoped to launch the next day.

Sunday

ukhasnet beach gateway

jcoxon and AI3

Meeting the next morning at seaside car park of Milford on Sea (after some last minute soldering in the boot of the car) we set out to walk along the shingle spit that extends out into the solent which has Hurst Castle at the end. The weather was mildly overcast and warm though the wind did pick up quite quickly. Midway down the spit we decided to setup the 3G gateway and then proceded to launch the nodes out to sea. It was quickly noted that the wind direction and waves meant that you couldn’t just place them in the sea but had to try to get them as far out as possible which involved throwing them. The large AI2 GSM Gateway didn’t do well in the waves due to its top heavy design and additional buoyancy so this was then relegated to the beach. AI3, PFLOT0 and MBb were launched out to sea a number of times but each time they floated back to the beach.It was also deemed too windy to fly the quadcopter.

ukhasnet AI1 on mast

We then moved further up the beach, walking past Hurst Castle and found a space on the end of the spit, re-setup the gateway and then launched the nodes, because of the change in tides and the angle of the waves the nodes didn’t come back but continued around the spit. We were able to track the nodes, PFLOT0 and MBb were sensor nodes and either were rx’d directly or via AI3 which was both a repeater and GPS node.
We were able to track the nodes over a period of an hour out to a range of 660m, following the progress of AI3 on the map. The limiting factor was that our laptop ran out of batteries, the pay and display was running out of time and it was lunchtime!

Conclusion

Apart from the weekend being a particularly fun day it was great to see that some of the technology and hardware we had developed was working. We successfully had a 3 floating nodes and 3 beach nodes communicating between each other over a range of over 0.5km (and we suspect we could have got even further with some planning). It was also good opportunity to take the hardware and then build enclosures that could stand up to the choppy waves and also the extreme launch techniques.

Plans

  • Try again, build upon current hardware. Personally plan to explore the further use of jam jars as enclosures as they sit low in the water which improves their stability but does risk limiting their radio range
  • Launch on a calmer day and be able to fly the quadcopter node to extend the range of the setup
  • More of the floating nodes to have GPS and repeating capabilities
  • Extend range by having floating gateways (GSM, APRS or Sat)
  • Deployment from a boat to avoid drifting back
  • Suncream, a picnic and more batteries

Links