Friday, July 12, 2019

Jun-Jul 2019: A Boat Monitoring Project

There is an old theorem in the yacht ownership universe ("yacht" being code for an over-complicated boat) that has several nuanced variations, but each version generally goes something like this:
  • Q: How do you become a millionaire? 
  • A: Start out as a billionaire and then buy a boat. 
So in that spirit, here’s another installment in what seems to be a continuing series of how to spend money on boating without (currently at least) doing much boating.

The Match-box Sized Trackimo Device Locates Ghost Rider on a 24-by-7
Basis via Cellular & GPS Triangulation & Until Now Has Been the Only
"Monitoring" Device Aboard the Boat.
As mentioned in a previous blog post, we had decided to pursue a remote boat monitoring solution for Ghost Rider.  Up until this year we had spent most of our time aboard, and any short separations were mostly when the boat was on the hard for maintenance. So this hadn’t been a priority until very recently when medical issues grounded us for a spell. While we had already installed a GPS locating device  with geo-fencing capabilities (Trackimo, which is similar to a LoJack device), and it gave us and the insurance company a certain level of comfort, more was needed.

Our objectives for a remote monitoring system had now gone beyond just knowing the boat still remained where we left it.  While there are other sensors we might add once we proved out the chosen system, initially we wanted to know about these high priority items when away from the boat: (1) house battery voltage; (2) shore power connection status; and (3) bilge water level.  Likewise, when any measurement varied from a defined allowable range, we wanted to be immediately notified via email or SMS text message so that intervention, troubleshooting and remediation could begin sooner rather than later.  

The “why” behind all of this is pretty straightforward – for instance, letting a house battery bank deplete to too low of a charge level (due to loss of shore power or a charger failure) could cost us around $5,000 USD for six new 8D AGM batteries.  Similarly, rising water levels in the bilge cavity could indicate seawater intrusion (say, from a failed thru-hull), something with which we have some nasty and expensive experience.  An early alert system of some sort would be much more cost effective than just crossing our fingers and hoping for the best.  Obtaining a truly decent ROI on any boat project is about as rare as an honest politician, so if we could actually find one then  it merited pursuit.
Our Initial Monnit Purchase: Two Remote Voltage Sensors, Plus the Cellular
Gateway Pictured with its Removable Antennae and Power Plug. After
Initial Testing We Added Two More Sensors.
Our system requirements included (1) a modular system where sensors could be added when/as needed; (2) a 4G LTE cellular gateway transmitter with its own SIM card to send notifications (U.S. based would be OK for now); (3) completely wireless sensors, preferably talking to the gateway in the 900 MHz radio frequency band for reliable wireless signal range; (4) a wide selection of said sensors for both current requirements and future expansion; (5) web-based and smart phone user interfaces for system administration and sensor reading access; (6)  a price point that provides a reasonable ROI….preferably under $1,000 (USD) for equipment outlay plus low-to-no subscription fees; and (7) a track record of quality and reliability, with decent online support.

Our 4G connectivity and wireless sensor requirements eliminated well over half of the potential field, leaving only a couple of contenders.  And a few more were priced in the “marine tradition” – i.e., if intended for a boat it must be unreasonably expensive.  We finally settled on a brand called Monnit™ (LINK), which oddly enough does not advertise its system for marine / vessel use, but boasts a fairly robust industry footprint among food services, data centers and general facility management.  We were also hopeful their client landscape would lend some staying power vs. some of the specialty (and over-priced) marine monitor systems that have flamed out or fallen behind the tech curve in recent years.
After Testing of the Voltage Sensors We Then Ordered a Water
Level Sensor for the Bilge

We were concerned most of all with the effective range of any water detection sensor in the bilge cavity, since that is deep within the bowels of the boat and closed off by a very well insulated engine room. Bench-testing the Monnit equipment back at the condo over 75’ distances and using closed metal closet doors had produced good results, and we weren't disappointed when it came to the real world install on the boat. The sensor transmission strength signals remained strong regardless of where we stuck the things.

Observations, Challenges & Concerns
Setting up the Monnit Gateway device was easy and straightforward, as was registering each of the individual wireless monitors so that they became part of the desired monitoring network. The same was true for defining allowable measurement ranges and monitoring intervals.  The browser-based configuration and reporting utility takes some getting used to, but with some practice it didn't take long to learn how to navigate and use it for system administration.  

Mounting and wiring each of the monitors was also fairly straightforward, although the one for the shore power tap required considerable sleuthing with the multimeter for a good tap-in location; and Rick had to attach (crimp and heat shrink) ring leads to the bare wire ends of the battery voltage monitor.  As an aside, monitoring for shore power presence wasn't exactly intuitive....see the "caveat" further down in this blog post.

The Yellow Arrow Points to the Installed House Battery Bank Sensor (in the
Lazarette) Which is Connected to the Positive & Negative Posts
 (Red & Black Arrows) of the Last Battery of the Six  in that Battery Bank.
The biggest concern going forward will be the sensor durability in the hot and humid marine environment in SW Florida, where salt-laden air could eventually affect any of the devices over time. The gateway won’t be of particular concern as it lives in the protected pilot house. The bilge sensor should fare OK as the sensor’s PCB and antenna are several feet above the actual water pickup sensor, and the engine room tends to be a hot, dry place anyway. The house batt and shore power sensors in the lazarette will likely have the most continuous exposure to the salt air, although their locations are protected from direct water sprays.  

As an aside, we corresponded with the Monnit engineers on applying a thin coat of CRC Corrosion Block to the sensors, but they weren’t crazy about that idea.  It also turns out they now have “Industrial” grade sensors with IP65 ratings which could likely last a good deal longer in the marine environment.  We’d recommend (as do the Monnit engineers) that more expensive option if just starting out, but see pricing comments below.   

It's also worth noting here that Monnit currently has two versions (what they call gen1 and gen2) of each sensor and you cannot mix and match them.  We went with the gen2 versions branded as "Alta" since they boast  improved circuitry and greater wireless range.   The other unknown is sensor battery life - these use a 3V coin battery advertised for an average of two years of longevity, but only time will tell.
The Shore Power Monitor (Yellow Arrow) Back
in the Lazarette, Where We Tapped into 
the Inverter (Red Arrow)

Conclusions
Overall our Monnit system has worked as desired and planned thus far.  We will want to figure out how to easily pause the monitoring network when it’s not needed (e.g., while cruising and living aboard) so that we don’t unnecessarily eat batteries, but that's a nit.  We’ll probably add some startup and shutdown actions to our existing departure and arrival checklists.  The web-based and smart phone remote access portals are also very good - see the screen captures at the end of the blog entry for some samples.

We also added a fourth sensor for the boat's cabin temperature, figuring it would be helpful to know if the air conditioning system takes a dive even if everything else is humming along nicely.  And for reference the whole thing still came in around $700 USD.  Using the aforementioned “Industrial” grade sensors, which probably (in hindsight) would have been a better idea, would add a premium to each sensor, so the total price tag would be closer to our magic $1,000 threshold, especially once tax and shipping fees are factored in.  Also note the industrial versions use a Tadiran lithium battery which cost four times what a coin cell battery does, but theoretically also last four times longer.

Some other notes on subscription fees and cellular communications:  With the Monnit system the standard sensor heartbeat interval is 120 minutes with sensor readings at 20 minute intervals (6 per heartbeat span) and typically should be sufficient.  There is a $10 per month data subscription fee for usage up to 1 MB, but there are no other extra fees unless you exceed  that 1 MB of gateway data ($5 for another MB, then $3 per MB thereafter.)  Generally that shouldn’t happen unless you let a sensor alarm keep triggering without intervention (which we discovered when testing the SMS and Email notifications.) 

If shorter / faster “heartbeat” intervals are desired then the “Premier” service subscription at an additional $39 per year may make more sense but is still a reasonable price; that also provides a host of other more advanced functions which may or may not be of interest.  As for the cellular portion, the chosen Monnit gateway uses an AT&T or Verizon SIM card, so it's fine for the U.S. where those carriers' coverage is decent.  Other options exist (Ethernet or custom SIM gateways) so we suspect the Monnit solution would work elsewhere.

The Temperature Sensor Tucked Away in the Port Aft
Corner of the Pilot House
A final procurement note:  for anyone interested in pursuing this solution, when you buy from Monnit you are buying direct from the manufacturer of the devices; that means when you order sensors they need some time (about 7 to 10 days in our experience) to actually build  and ship them….they don’t carry inventory.  That said, their online (email) and telephone support is excellent.

There is another and nearly identical purchasing option at the "My BoatStatus" web site (see links below), a company who also source their devices from Monnit; they do carry inventory, but there you will find some price markups vs. the Monnit direct pricing.  Other brands that were considered are listed near the bottom of this blog entry along with their associated web site links.

We’ll provide periodic updates on the system’s reliability and durability, but for now we are quite pleased with the result, and are calling the project a success.

Caveats on Shore Power Monitoring
Wiring up this particular voltage monitor was a challenge, as how and where to monitor the AC voltage coming in from the shore power pedestal isn’t exactly obvious...at least it wasn't to Rick.  This is partially due to Rick's grade school level of electrical expertise, but also to how the boat’s charger/inverter unit works. 

When our (Xantrex) inverter is online and senses the presence of shore power current it simply goes into a pass-through mode and sends that power current to the boat’s AC distribution panel.  But if/when shore power is interrupted it will automatically and immediately go into its inverter mode and start generating ("inverting") 120 VAC power from the 12 volt house battery bank, and pass that power to the AC panel.  So monitoring for the presence of AC power downstream of the inverter unit is not particularly beneficial for this purpose.  (You can put the inverter in “bypass” mode but you would have to accept that a loss of shore power would result in an electrically dead boat….that would be OK if you planned for an extended layup but otherwise would be problematic.) 
The (Bilge) Water Sensor....Red Arrow Points to the Actual
Water Sensor Pickup (Mounted on the Shelf for the High Water
Bilge Pump).  Yellow Arrow Points to the Sensor Housing
and its Antenna Mounted a Couple Feet Above the Pickup.
This is in the Engine Room Just Aft of the Main Engine,

One could also place the voltage monitor tap at a constantly powered 240 VAC circuit or outlet (the inverter only provides 120 VAC, the 240 VAC juice has to come from either shore power or the generator), but that’s not as easy as it sounds, at least not on Ghost Rider.  The other alternative we eventually figured out was to just tap into the AC lines coming into the inverter, which is what we ultimately did.  That tap is only on one of the two incoming 120V shore power legs, but that's sufficient for our purposes.

Other Competitors Considered
Other solutions considered but not pursued further include those listed below.  There are probably others out there.  Several of these also include the GPS location and geo-fencing service within their monitoring solution, which we already had installed via Trackimo and did not want to replicate:

➢ Boat Command….requires running wired cabling (LINK)
 Boat Nanny….acquired by GOST
 Go Free marine…. requires running wired cabling (LINK)
 GOST….way too pricey for us (LINK)
 Kobelt….very sophisticated, requires N2K, could not find pricing (LINK)
 Maretron….requires N2K wire runs, also quite pricey* (LINK)
 Mazu Marine….satellite based, hefty subscription, but works anywhere (LINK)
 My Boat Status….uses Monnit-supplied devices, too, but with price markups (LINK)
➢ Simrad BoatConnect....also requires running wired cabling (LINK)
 Siren Marine….their wireless version is fairly new, good reviews, but pricier (LINK)

(* If you're already a Maretron user this "pricey" comment may not apply and would be worth evaluating further.)

What’s Next
Oh yes indeed, Rick already has another boat project in mind….this one to find a CCTV camera solution for visual monitoring of certain conditions / areas.  Once again the desire is to find a mostly wireless solution to avoid tearing the boat apart to run cabling.  It should be an interesting exercise.

Some Screen Shots
Here are some application screen shots, both from the iMonnit web portal as well as the included smart phone app.  (Side note: their web portal sports two user interface experiences....Rick tends to favor the older "classic" look, but that's highly subjective.)  At any rate, these are just a small subset of what's available:
Screen Shot of the List of Notifications (Event Triggers) Defined for the Four Sensors and
Gateway Device...from the iMonnit Web Portal

Screen Shot of Shore Power Sensor Readings from the iMonnit Web Portal
Screen Shot of House Battery Bank Sensor Readings from the iMonnit Web Portal
Another Screen Shot Demonstrating the Graph Depiction of the House Battery Readings...That Spike is During a
"Bulk Charge" Period.  And Yes, These Are Obviously Temperature-Compensated Voltage Readings.  It's HOT Here.

An Android Smart Phone App Screen Shot.....This Example
Showing House Battery Bank Sensor Readings
Another Android Smart Phone App Screen Shot.....This One Showing Graph & Sensor 
Readings for the Shore Power Monitor

1 comment:

  1. POSTSCRIPT FROM RICK (MID-AUGUST 2019):

    We did encounter some problems with the (bilge) water sensor. After it got submerged during our annual bilge pump testing routine the thing would not dry out....it kept alarming as "water detected" even after drying it out with a hair dryer. Monnit issued a RMA and they shipped out a new one. Rick fashioned an L-shaped bracket out of starboard so this one would hang vertically and several inches below the high water bilge sensor -- it should shed water better that way plus that provides an intermediate warning level between the low & high water bilge pump float switches. That replacement sensor has been performing per spec. It took a while, but Monnit tech support was fair and attentive throughout.

    ReplyDelete