Showing posts with label industrial iot. Show all posts
Showing posts with label industrial iot. Show all posts

Wednesday, November 11, 2015

Monitor access to buildings

Access to barns and recording the access is more important today than ever before.  Biosecurity and monitoring labor are just two of the reasons it makes sense to have a basic understanding of the coming and goings of people at the farm site.  I have put together using a basic door sensor a system that can provide quite a bit of insight into a farm's activity and compare to other similar farm's activity for benchmarking purposes.

Certainly I realize that this is a fairly low-cost monitoring system.  Undeniably there are bigger and better ways to monitor exactly who is coming and going and at exactly what time.  You could retina or fingerprint scan workers like 007 movies but at what cost?  You could have card scanners but then look at the expense of installing that for one or two workers.  What if they forget their card or lose it?  Often the "home office" is far away and the logistics of getting a new card could take days.  Or a substitute needs to go to the site who doesn't have a card?  A simple low-cost employee monitoring device could be the solution.

Here is what I did.

I installed a simple door sensor that sends a count of how many seconds the door is open in 5 minute intervals for each time it is opened.  That information is then sent to Grovestreams where I am doing some simple analysis with it that provides me with some excellent information.

Data #1.  I get the first time of the day that the door was opened up.  This is huge.  If the employee is supposed to be at the site at 6am and they regularly arrive at 9am that is a problem.  


The image above shows activity on a system that monitors a door at my office.  Grovestreams allows for an easy way to determine the first and last timestamp of activity received for the day and shows it in a dashboard.

In addition I can take the last timestamp and subtract the first and get an estimate of the time the site was occupied for the day.  

I also chart this for quick viewing (guess I haven't been to the office at all in the last few days):


Interesting alerts on this.

1.  Send an alert to email or phone on the first arrival every day.

2.  Send an alert to email or phone on activity outside of a time filtered period.  Example send an alert for every activity outside the hours of 6am to 6pm.

3.  Send an email at 12:01am if the duration of the previous day was less than 4 hours.

4.  Same alert as #3 except for exclude weekends.


Expanded analysis:

Many sites have multiple barns.  As part of the Multisense sensor gateway users could add door sensors to each barn entrance and managers could track how much time is spent in each of the barns on the site.  Alerts for barn doors left open would be simple to include as part of any package as well.

Grovestreams and Multisense make prototyping IoT projects simple and low cost.  This door sensor project took me hardly any time or money and now I think this could be used in barns all around to monitor employee time.

Thursday, September 10, 2015

Grovestreams is a great IoT platform

Grovestreams is an IoT platform that is under the radar now; but that shouldn't last for too long.  I am listing 10 reasons Grovestreams is a great IoT platform for our solution.  Farmstreams is built on the Grovestreams platform and is leveraging the tools built to deliver value to livestock production.

1.  Simple to put data in and take data out of the store.  The APIs are published and open.  Easy enough that non-programmers can accomplish it.  Powerful enough for talented software developers to build apps and websites pulling and putting data in and out.  This also allows larger organizations to leverage Grovestreams for its cloud-based data gathering and analytics without sacrificing or needing to switch from the enterprise software they run their business with.

2.  Built with Big Data technology.  Scales to nearly unlimited size.  When it needs to get bigger, a new cluster is added.

3.  Cost.  A user can get started with a Free account to play around with developing a solution.  Pricing is data driven.  You pay for what you use.  No mandatory consulting fees or annual contracts.

4.  Quick to deploy a solution.  Whether its an app or a sensor feed, its easy and fast to run calculations, build dashboards and deploy your IoT solution.  Location/Map integration is also a nice feature.

5.  The roll-up stream data is a really efficient way to get to the data you need.  The in the field sensors can simply send raw data in.  Once received, they go to the store but then are accessed with the roll-up time selection (example hourly) and the preferred statistic (example average).  So, if you want the average hourly humidity, send the raw data in every minute and the data automatically rolls up to the hourly average.  Furthermore, minimum and maximum are also readily available in the store - so you can also view average, minimum and maximum on the same chart.  Even furthermore, you can monitor the quality of your data by monitoring gap counts in the rollup streams.

6.  The analytics package is really good.  You can run very complex expressions on any stream in the store.  You can then run more expressions on those created streams.  For example, I am calculating a feed bin inventory using data inputs from 5 different sensor streams and/or app entry streams.

7.    Blueprinting an organization makes setting up new organizations really easy.  When new sites (like a farm) needs to be set up simply create a blueprint and start the new organization exactly as the previous one was very quickly.  Also, switching organizations is all done in one log-in so a user can have access to all of their organizations without having to log out and log back in.

8.  The alerting engine is solid.  Users can get alerts upon arrival for those alarm type streams.  Users can run a condition alert, example temperature  > 32 send alert.  Or with latency, too much time has passed between data streams.  Features like time filters are effective in mitigating false alerts.  Example: feed latency alerts are ignored in the middle of the night.

9.  The dashboards to visualize the data are really good and easy to use.  Drag and drop graphs or spend more time and develop a super-dashboard.  A lot of well thought out features to quickly and accurately create the visual tools needed for stakeholders to your project.

10.  We are just at the beginning.  The base platform is already great.  With the enhancement roadmap laid out the platform will get even better.

So, why would I boast about this platform?  Shouldn't this be my best kept secret?  First, I want customer farms and potential customer farms to know how good this back-end solution is.  Feedlogic's focus is on providing intelligent and connected devices at the farm.  Without a platform like Grovestreams that endeavor is somewhat pointless,  Second, the IoT space is like a really large sandbox right now.  There is plenty of room for a lot of people to play in it.  For Farmstreams to continue improving more and more people need to find Grovestreams and use it.  Then Grovestreams can grow its support system, add enhancements and developer networks can be started for collaboration within and across industry can make all solutions better.  I encourage you to check it out or contact us with questions/comments.