ag demo FAQ
Where does nio interact with data in this system?
At Deep Sky Vineyard, we feature nio within a distributed architecture that has nio instances interacting at every level of the system from the visualization down to the small sensor data aggregation nodes.
How much sensor data is generated?
At Deep Sky Vineyard, approximately 4 Megabytes of raw data is generated per day. This includes data from physical sensors, data generated by our software, system metrics, and user generated data. If this data were collected without n.io filtering, it would add up to roughly 1.5 Gigabytes per year for our 20 acre demo.
Will this system work in a remote area with limited bandwidth?
The nioAg platform is optimized to run locally using very little bandwidth of the cellular uplink. Currently we run this system on a 3Mbit Up/10Mbit Down connection and the bandwidth required to run nio is only 10–15 kbps at any given time.
How much power does the system use?
To continuously run our entire system as installed, it requires less than 500 Watts of power. Under normal full load conditions, we’re using 450 Watts to power our system on a 110VAC power feed. This measurement does not include the sensor extensions, which are battery powered and recharge via solar power.
Why are there so many computers?
Keep in mind that the nio edge nodes are single core “computers.” The nio Ag demo at Deep Sky Vineyard was designed with much larger installations in mind, so high expansibility is designed into this initial pilot. Because the nio platform easily operates in a distributed manner and we wanted to validate the Deep Sky model and its ability to scale to larger operations, we decided to go with a more advanced distributed layout versus a simple star configuration with all fields directly reporting to the shed.
Can I use my existing sensors?
With interoperability at the heart of n.io, we’ve done our due diligence to ensure that many of the commonly available agricultural sensors will be able to connect to our platform. If there is a sensor that doesn’t currently connect, we’ve incorporated up to 50% expansibility overhead into our sensor hubs and extensions to allow for new connections and software blocks that may be needed at the edge.
What is the system's estimated lifespan?
Starting from the smallest elements of our system—we estimate the sensor extensions are able to run unassisted for 3 years, which is only limited by each battery’s lifespan. Moving up to the larger elements—we foresee life spans of 15 years or more with nominal maintenance.
How is maintenance performed on this system?
nio is used to notify the farmer if a sensor has gone bad or if an individual component in the system is performing below expectations. Hardware used in our platform consists mainly of Commercial off the Shelf (COTS) items that can be replaced by the end user. Software configuration management is performed remotely to allow for high system availability with maximum uptime and minimal worry for the farmer. Local notifications and management can be used if remote access is not available.
Are there any standard requirements on the sensors?
nio does not require sensors to use specific standards. Our solution is hardware independent. nio supports multiple communications protocols and can be easily extended to support more.
How do the sensors hookup to the nio network?
The sensor nodes at Deep Sky Vineyard are solar powered and communicate with a base station using the IEEE 802.15.4 wireless protocol. This solution works well for us, but nio can use many different standards to communicate.
I am assuming you need cellular network coverage to channel all of the data generated on the farm to the cloud. Is there any technology where you could support farms that don’t have cellular coverage?
One of the advantages of nio is that all data stream processing can occur at the edge or at local higher-tier nodes. At Deep Sky Vineyard, we only use the cloud for analytics of long term trends and data retention and for remote visualization and control. For a farm that has no cellular coverage, any other suitable communication technology can be substituted or data can be collected and visualization and control can be made available on site.
How is n.io monitoring the well pump motor?
Within our industrial client environment, we are monitoring temperature, vibration, and amperage sensors on all mechanical components to predict failure. At Deep Sky Vineyard today, the motor amperage is the only mechanical metric that we are currently sensing.
How much new equipment did Phil have to buy to integrate nio into his vineyard?
We integrated Deep Sky’s Davis weather station and all of their existing irrigation infrastructure, including the pumps, filters, chem tank, and valves at each block. As far as the in-field sensors, Deep Sky did not have any of those components in place to start with, so all of those items were integrated as part of the project.
Is nio just the software/monitoring platform, or are you also making/supplying the nodes/controllers also?
Due to small scope of this pilot project, our team chose to build, install, and integrate all of the necessary hardware.
Predicting pump failure is important. I realize you can check amperage, temperature, and vibration, but how do you predict possible gasket failure or stress on the pump components?
Amperage, temperature, and vibration are the ‘holy trifecta’ of predictive machine failure. nio’s ability to process this trifecta on a lightweight and affordable basis creates feasibility that generally is not available on a per asset basis. Specifically, gasket, bearing, or other component wear or failure will be detectable by the trifecta.
How many sensors are typically installed per zone and what is typical zone size?
The actual configuration is entirely up to the farmer. The average zone size at Deep Sky is 1.2 acres. We currently have 1 wireless sensor extension (3 sensors) in all zones except for one, where we wanted to demonstrate the “art of the possible.” In the 12th zone, we installed 3 sensor extensions with 19 total sensors to get a full context picture of the conditions throughout that zone. This zone includes 3 soil moisture probes, 9 tensiometers, 3 fruit zone temperature and relative humidity sensors, 2 dendrometers, and 2 leaf wetness sensors.
Who owns the business process logic or algorithms developed and applied to Deep Sky’s requirements? How easy is it to modify this model for other ag activities or even unrelated business verticals?
Although important input was received from Deep Sky’s owners and various vineyard experts, nio was responsible for the business process and algorithm development. nio is built to be highly flexible and capable of easy configuration to meet the demands of the particular use case, regardless of the vertical.
Does the customer also get an alert if a sensor is not transmitting correctly, or goes offline for some reason? If so what does that look like?
Yes, nio is self aware and configurable to send alert messages to the farmer if something is wrong with the system. (Example: The soil moisture sensor has not reported data in X amount of time.) These alerts are no different than the n.io alerts demonstrated in the webcast and can be easily configured to the user’s preference.
Is nio able to measure rainfall and add or implement those measurements for future watering schedules?
Yes. nio can measure actual rainfall using the Davis weather station at Deep Sky Vineyard. The way the irrigation algorithm is configured, rain is integrated into the calculations just like the drip irrigation and affects the model in the same way. The variable driving the model for when and how much water to apply is the soil moisture per zone vs. the farmer’s high and low thresholds.
Several research groups have shown dendrometers are not a reproducible and reliable way to measure day to day variation in water potential, but rather more appropriate for long tem scales (months to years). Why use dendrometers to measure stem water potential with n.io?
We agree that dendrometer data is much more valuable when viewed over a long period of time to understand various trends. The typical way to measure stem water potential is through a very manual process using a pressure chamber. In an effort to make the overall system as automated and seamless as possible from the farmer’s perspective, we chose to include the dendrometer in the initial pilot solution in an effort to correlate the data and better understand its viability for future solutions. It has not yet been determined if dendrometers will be included in future vineyard projects.
Is the annotation screen just using the standard gps from a mobile device?
Yes. In the case of this pilot project, we tested on both iphone and android devices using their standard built-in GPS sensors.
How do you get accuracy that closely? By row?
The accuracy that we can expect from most smart mobile devices today is within 3 meters…and getting better. Because the rows at Deep Sky are 10 feet apart (3.048 meters), we are able to determine the exact row on the vineyard. If nio is unsure which row the farmer is in, it gives him/her the option to choose between the 2 closest.
What is the cost to own a copy of nio? What are the mechanical installation requirements to enable nio, such as: probes, sensors, weather stations, etc.?
We elect not to share nio licensing pricing via Q&A. There are no mechanical restrictions on hardware with nio at this time. If you can get a signal from a sensor/data source/device, nio can be configured to incorporate that into an ecosytem.
You mentioned that nio is a patented AI solution—are there any plans in the future for potentially shifting a portion or all of nio to an open source platform?
There are plans to release elements of nio to the open source community in the future—stay tuned!
Who handles sensor installation/selection on the farm?
We worked with Deep Sky to select the desired sensors for the project. All of the build, install, and implementation was handled by the nio team. In the future, we will enable the farmers or integrators to handle field requirements.
How specific and limited is the patent you speak of? Does it relate only to nio’s general capabilities, or does it relate specifically to nio’s application within agriculture?
We have pending applications and issued patents that address n.io’s general architecture and capabilities as well as specific use cases. Of these, several are directed to n.io’s capabilities relating to agriculture.
I would like to know more about the system architecture behind the scenes, including how to collect the data and how to deploy the gateway and the sensors/actuators?
We employed a multi-tiered distributed architecture at Deep Sky. We will expand on this during the upcoming technical deep dive.
If your platform goes down, is there a backup system?
Redundancy is built into nio. Because the data is processed at the edge, if any edge node were to go down, other nodes can take on the load. No one node failing can take down the whole system.
What sensors are needed for soil moisture and how do you measure per plant?
We use COTS soil moisture probes that contain 6 soil moisture and temp sensors down to a 40″ depth. For the purposes of the pilot, 11 of the 12 zones have one probe while the 12th zone contains 3 probes and 9 tensiometers to determine soil moisture and irrigation decisions.