JavaScript is off. Many parts of this site require JavaScript to function properly.

Collecting Data Through the Canary API

10:23 AM

Bringing data into a historian and then later attempting to access that data and move it to another system can sometimes be a challenge.  The Canary Logger is designed to connect to your data sources through an OPC server and then delivers data to several clients through web services.

However, if you would like to do more or connect in other manners, we have made it easy by offering a very easy to use API.

Canary Sender API

The purpose of the “Store & Forward” application is to send data from remote locations to a secondary machine that will then push the data into the Canary Enterprise Historian. The application consists of two web services that communicate with each other through a specified port. The names of the two services are the “Sender Service” and the “Receiver Service”.

Sender Service
The “Sender Service” is a .NET WCF application. An application will use this service to send data to the Historian to be stored. Best practice is to place the “Sender Service” on the machine with the logging application. This service will buffer data that is passed into it and forward it to a specified “Receiver Service” when the connection is available. There are fail safes in place to protect the data when connections are lost. The data will not be removed from the buffer until it receives a response from the “Receiver Service” confirming that it can release the data.
Any web service compatible client can communicate with the service using the http protocol. The default port that the “Sender Service” uses is port 55251. A different port can be specified in the “Sender Service” SAF_SenderService.exe.config file if necessary. In addition, a .NET WCF application running on the same machine as the “Sender Service” can communicate with the service using the net.pipe protocol.
An html page that shows the “Sender Service” status can be viewed on the “Sender Service” local machine’s browser at http://localhost:55251/saf/sender/status. This page will display the number of clients hooked up to the “Sender Service” and information specific to each session. Information includes the session id, logging status, initial “Receiver Service” connection time, number of tags being logged, number of tvqs in buffer, number of properties in buffer, number of annotations in buffer, number of errors, and the last error message that occurred.

Receiver Service
The “Receiver Service” is used by the “Sender Service” to put data into the Canary Enterprise Historian. A client application cannot use the “Receiver Service” directly.
The “Receiver Service” must be located on a machine that also has the Canary Enterprise Historian software installed on it. This is necessary because the “Receiver Service” pushes data into the Historian through a local interface. When data has been successfully stored in the Historian a response will be sent back to the “Sender Service” to clear the data from the “Sender Service” buffer.
The default port that the “Receiver Service” uses is port 55254 for http and 55256 for net.tcp. A different port can be specified in the “Receiver Service” SAF_ReceiverService.exe.config file if necessary. The “Sender Service” always uses the net.tcp to communicate with the “Receiver Service”.
An html page that shows the “Receiver Service” status can be viewed on the “Receiver Service” local machine’s browser at http://localhost:55254/saf/receiver/status. This page will display the number of clients hooked up to the “Receiver Service” and information specific to each session. Information includes the session id, initial “Sender Service” connection time, number of tags being logged, number of tvqs that have been written to the historian, number of properties that have been written to the historian, number of annotations that have been written historian, number of errors, and the last error message that occurred.

Helper Class
A .NET helper class has been created for the “Store & Forward” application. The purpose of the helper class is to provide an additional interface for .NET applications that can simplify use of the “Store & Forward” technology. It can help simplify the connection process and provides a variety of overloads for each of the interface methods.

Transform Equation
Transform equations can be used to transform values as they are being logged. Each tag can be configured with its own transform equation string. The transform calculations use a library called NCalc to parse the equation string provided where the keyword “Value” is replaced with the passed in tvq value and the calculation is then executed. As an example, the transform equation string “Value * 1000” will multiply each value for the specified tag by 1000.

Time Extension
Time extension is used to extend the time for the last value of a tag if the tag is still logging but a new value has not been stored for an extended amount of time. If a time extension for a tag is set to 15 seconds and the last value has not changed in the last 15 seconds the value will be updated with a current timestamp. If a tag value is then stored with a timestamp that occurred before the extended timestamp but after the timestamp at the start of the last value, the extended value will be modified to reflect the corrected timestamp.

Buffer Files
Default buffer files will be placed on the “Sender Service” machine in the path “ProgramData/Canary Labs/Logger/StoreAndForward/SenderService/{historian}/{client id}” where the {historian} and {client id} are the historian and client ids that get passed to the “Sender Service” when the session is initialized through the GetSessionId method request. Buffer files get cleaned up automatically, but the paths to the files will be left intact.

Error Logs

An error log containing all “Sender Service” and “Receiver Service” errors will be created on each machine. The default error log file path will be “ProgramData/Canary Labs/Logger/StoreAndForward/{service}/{service}.log” where {service} will be either “SenderService” or “ReceiverService”.
Read More

Enterprise Historian Data Sheet

5:39 AM


Founded in 1985, Canary specializes in centralizing, distributing, and analyzing time-series process data.  The Canary Historian is a proprietary, non-SQL database that can gather millions of data points without data loss or purging.  
With over 17,000 installations around the world, Canary’s solution can be implemented as a small local historian or scaled to a full enterprise level solution, allowing for both a
simple and clean installation as well as on going database administration.
Custom API and ODBC connections allow for easy data flow and a variety of web service clients will ensure the entire organization can access the historical process data when necessary.

Data Acquisition

Canary has many ways to collect process data.  The most common form is via OPC server, allowing for direct connections to PLC, SCADA, HMI, or sensor data.  The Canary Logger Administrator connects directly to the OPC server and obtains tag names and common profile information from the OPC server.  You can configure your data as needed, changing the tag name, description, properties, or dead band.  You can also perform a single tag data calculation in the logger.  Within a logging session you can create logging groups.  Each logging group can be associated with an individual data set, a collection of specific tags that are grouped in the historian for better tag organization.  
The logging group can be customized to have preset update rates so that you are not forced to accept the value change rate that is delivered from the PLC/SCADA system.  You also have the option to create a logging trigger, allowing you to start and stop logging a group based on a specific tag’s value.
The logger passes data through to the historian using the Canary Store and Forward Service.  Store and Forward is comprised of two components, a Sender Service and a Receiver Service.  These two services communicate using Windows Communication Foundation (WCF) and all data is encrypted during communication.
The OPC server, Canary Logger, and Store and Forward Service can be installed on the same machine as the Canary Enterprise Historian or sit on their own independent machine and connect to several historians across multiple networks.  If contact is lost between the Sender and Receiver Service, the Sender Service will cache data to local disk.  When communications return, the cached data is transferred to the historian in time sequence order and removed from the Sender Service.
Not only can one Sender Service connect to more than one historian, but multiple logging machines can also be configured and networked to one historian.  These loggers can be setup across multiple networks to monitor remote sites as well.  The Canary system does not limit the number of loggers used, and there is no additional charge for adding extra loggers as needed. 
The Sender Service can connect to SQL databases using the Canary SQL Connector and a custom API is in place as well.  Using the API, you can write your own data into the historian and eliminate the need for both the OPC server and Logger program.  The API also supports dynamic tag configuration.  In addition to the SQL and API Connector, the Sender Service can import CSV and flat files.

Historian Database

The Canary Historian is a non-SQL, proprietary database that specializes in the storage and retrieval of time-series process data using “lossless” compression.  Unlike other historians, Canary never converts actual data to approximations.  Also, the historian never purges or dumps data and resolution remains the same whether the data is one day or several decades old. 
The database is comprised of self-contained .HDB files and features time-based organization.  Every data update contains a Time Stamp, Value, and Quality (TVQ) reading.  Generally, unless otherwise specified by the user, files are rolled over nightly, allowing for simple archiving and data retrieval.  Files can be taken offline and easily moved from one historian to another making system restoration extremely simple.
The historian is capable of write speeds up to 2,800,000 TVQs per second and can support sub-second data with a timestamp resolution of 100 nanoseconds.  Manual data entry is also supported and can be done directly into the historian.  Read retrieval speeds approach 4,600,000 TVQs per second and due to the time-based database structure allow for effortless data recall even when spanning large timeframes.
An HDA server is available for third-party consumption as well as the Canary Trend Reporter client.  
The Canary Views Service connects to the Historian Service and allows for client interaction via WCF.  Canary’s Axiom Core, TopView Alarming client, and Excel Add-In client connect to the historian using the View Service.  Also, the View Service allows for both a custom API and an ODBC connectors to be made available for data retrieval.  
The Views Service will also allow for connections to multiple other Canary Historians through the Canary Mirror Service as well as third party historians using either HDA or UA.  All data transferred between the View Service and other clients is encrypted and passes through a single firewall port.

Canary Administrator

The administrator tool makes data historian admin simple and quick. Using the Canary Administrator, you can manage multiple historians and sites, quickly toggling between them using a simple drop down selector. 
From the administrator, you can monitor the logger, sender, receiver, and historian.  Visual indicators allow for quick notification if one of the services begin to buffer as well.  Email notifications can also be setup to notify users if any data sets stop logging.
You can easily view, stop, and start all services as need be as well as monitor the number of clients that are connected through the View Service.  A messages tile allows you to view any warning or error messages, as well as view a full audit of all past messages.
A License Administration tile will make the monitoring, transfer, and configuration of available licenses convenient.  It also allows for the addition of licenses through the internet as well as manually entered.  
The Historian tile allows you to view each individual data set and configure each one individually.  From there, you can set up email notification parameters if a data set stops receiving data as well as schedule rollover frequency.  If you would like to purge data after a set period, this can also be scheduled from the Historian configuration tile.  
Data sets can be closely monitored and raw data can be quickly obtained as well.  Since the database is sorted by time, you can quickly scan through years of data to find an exact period.  Then, select the tag and have access to all of the raw data as well as the tag properties.  A simple trend tool allows you to visualize the last 1,000 TVQ values without having to leave the administrator.    

Security and Access

Security remains as high a priority to Canary as our data integrity.  Our focus is on both secure client access as well as data logging security.
Canary Administrator access is restricted to only users or groups designated by the
administrator.  This privilege allows for complete monitoring and management of the Canary system, including the monitoring of active client connections and data access services.  Also included is the ability to audit all manual data entry as well as original data overwrites.
To guarantee secure data, access to the Canary platform is limited using a Microsoft Windows security infrastructure.  An administrator can use local security or active directory user accounts and groups.
Using the above methodology, the administrator can control access to not just historians and data sets, but individual tags as well.  
With multiple data logging options available, local and remote access is restricted to a single port per service.  Port numbers may be changed from their defaults and communications can be encrypted using either a self-signed certificate or a certificate issued by a trusted root certificate authority.

Data Mirroring and Redundancy

The Canary Historian provides the mirroring of stored data on multiple historians to provide high levels of data redundancy as well as to simplify data retrieval.  However, historian redundancy is not limited to pulling data through the Canary Mirror Service.  Since each individual Store and Forward Service can point to multiple historians, systems can be developed that push data to multiple databases.  This model allows for both communication and database redundancy including fast disaster recovery.  
With appropriate security privileges, data reads can be made across any of the connected historians.  Users can have access to multiple data sets and tag values across multiple sites while still being restricted to other historians or tag groups.  

System Requirements

Canary Enterprise Historians can store tens of millions of tags and can support timestamp resolution to 100 nanoseconds.  The system requirements below are based on a total system size of 1,000,000 tags with up to 300,000 tags at each local historian.  Core count, processor speed, and hard drive space may need to increase as the system increases.  It is assumed that each data tag has one second resolution with a sixty percent change rate.  

OPC and Logging Server   Store and Forward Server
  • Dual Core 2.0 GHz Processor
  • 8 GB of RAM
  • Windows 7 64 bit or greater
  • .NET 4.5 or greater
  • 500 GB HDD @7200 RPM
Local Historian Server
  • Quad Core 2.4 GHz Processor
  • 16 GB of RAM
  • Windows 7 64 bit or greater
  • .NET 4.5 or greater
  • 1 TB HDD @7200 RPM
Corporate Historian Server
  • Six Core 2.4 GHz Processor
  • 32 GB of RAM
  • Windows Server 2012 or greater
  • .NET 4.5 or greater
  • 2 TB HDD @7200 RPM

Read More

The 5 Trend Features You Must Have

5:26 AM

Canary Software’s Five Must Have Trending Features

For over 30 years Canary has specialized in providing fast and easy to use trending tools.  Below are five features that continually receive praise from Canary users.

Display More Than 200 Trends

Your screen resolution is the only restriction to the number of trends you can load.  Canary allows you to full or auto-band your trends so you can decide how your screen should be designed for optimal trending performance.

Define Limits and Thresholds in Seconds

In less than thirty seconds you can create high and low limits for each trend, add multiple limit lines to frame operating thresholds, and even shade the trend when the limit is reached.  Canary provides a simple interface that gives you full control of features like line type, color, and even thickness.  Beyond basic limits, you can even apply one trend as a limit for another!

Perform Calculations

Need to calculate pump efficiencies or power factor ratios on the fly?  Use Canary’s calculated trend feature and create complex or simple calculations involving multiple tags.  Choose from more than 20 mathematical operators such as SIN, TAN, SQRT, COS, IF, MIN, MAX, and more.  Once the formula has been defined it immediately calculates a new trend, even providing historical data and the ability to export the newly calculated data.

Time-shift and Aggregate Data

Easily compare current data to historic data with time-shifting.  Canary allows for the same tag to be loaded multiple times on a single chart.  Once loaded, you can quickly time-shift one trend and band it with itself to create a custom graph showing live data overlaid with historic data .  Add to this feature the ability to aggregate data based on time interval, and you can process data into Max/Min, Range, Total, Count, TimeAverage, or one of over thirty-five options.

Export Directly to Excel

Want to move raw or processed data quickly into Microsoft Excel?  You can export any trend chart directly into Excel using Canary’s built in export tool.  Just select the tags and define the time-period, and Canary automatically launches an Excel workbook with the requested data.  Or, bypass the trending tool completely and import raw and processed data directly from the historian into Excel using Canary’s Excel Add-in.  Built right into the Excel ribbon bar, the Excel Add-in allows you to search any historian on the network and import the data directly into the workbook.

Many Other Features

Contact Canary today and learn about the many other features offered, including annotations, chart sharing and permissions, and custom dashboards.

Read More

Canary, a Proud SyTech Strategic Partner

12:59 PM

SyTech, Inc., the leading provider of automated report generation and data analysis software for industry and Canary have become strategic partners, providing seamless integration between the XLReporter and Canary Enterprise Historian.
The Canary Enterprise Historian and Axiom applications are used in complex environments and easily handle in excess of 1 million points per second while producing over 200 trends per chart, enabling organizations to quickly and effectively analyze data without delay.
XLReporter has been bundled in the Enterprise Historian as TrendReporter since 2008. TrendReporter manages raw data from the Enterprise Historian into meaningful information by applying powerful analysis routines.
"Our customers have been receiving great benefit from the XLReporter. It is easy to use, integrates seamlessly with Enterprise Historian, does not require any programming and is based on the familiar environment of Excel," says Ed Stern, Vice President of Canary Labs. "Going into Strategic Partnership with SyTech reinforces our offering to customers that require reporting as a part of their overall automation system."
"The partnership of Canary Labs and SyTech is a solid endorsement of the value XLReporter brings to Canary Labs Enterprise Historian," states Peter Kaprielian, CTO of SyTech. "Having been fully integrated into Enterprise Historian, we made sure that our joint users benefit from the best of both worlds. Our commitment has been set in cement."
XLReporter is fully compatible with Microsoft Windows and Windows Server operating systems. Support for Excel extends to Office 2013 on both 32 and 64 bit platforms.

About SyTech

Founded in 1991, SyTech develops software for report generation and analysis. The product, XLReporter, is the recognized leader in industry for automating and distributing reports. XLReporter delivers connectivity to all the major HMI/SCADA, Historian and business data sources through proprietary or industry standards. The product is sold by automation software distributors and OEMs with implementations in over 80 countries around the world. SyTech markets include manufacturing, utilities such as water and wastewater, building automation and energy management. Their award-winning software is endorsed by small municipal facilities as well as major international manufacturers and Fortune 500 companies.  Learn more by visiting

About Canary Labs

Canary provides Enterprise Historian and trending solutions that simplify and optimize data analysis driving more informed, confident decisions. Leading companies worldwide rely on Canary's open, flexible and high performance software to improve process metrics and increase the agility, efficiency and reliability of data access. With a reputation for stellar service and support and minimal cost of ownership, Canary ensures customer success to a client base with over 17,000 installations in 36 countries.
Read More

The Advantage of Smart Building Monitoring

9:05 AM

Smart buildings integrate major building systems onto a common network and share information and functionality between systems. Monitoring these systems will not only improve building operations but also lower overall operating costs. Thus, the implementation of smart building technologies will improve energy efficiency, enhance operational effectiveness, and most importantly, will increase tenant satisfaction.

Using Canary software, customers can maximize energy efficiency, greatly reduce unexpected equipment failure, and transform the facility management process while implementing reductions in material use. For example, Energy Star reports that a simple one degree temperature adjustment can save four to five percent in overall energy costs.

Real-world Application

At the core of a smart building is the Canary system. Providing centralized visibility of a building’s components such as heating, ventilation, air conditioning, lighting and other systems. Further examples of real-world applications include:

Air Handlers - Responsible for mixing and returning outside air to reduce temperature and humidity conditioning. Expenses are reduced by using less chilled or heated water. However, to keep the building’s air healthy, some external air is required. To optimize energy efficiency while maintaining healthy indoor air quality, the amount of outside air ventilation can be adjusted based on the measured levels of occupancy.

Chilled Water System - Primarily used to cool a building’s equipment as well as the internal air. This includes chillers and pumps. The chilled water system has analog temperature sensors that measure the chilled water supply and return lines. The chillers are sequenced on and off to condition the chilled water supply. The most important component of good chiller control is understanding the full and partial load requirements, and then properly matching those to system performance.

Hot Water Systems - Supplies heat to the building’s air-handling unit. Includes boilers and pumps. Analog temperature sensors are placed in the hot water supply and return lines. A mixing valve is used to adjust the heating water temperature loop. Boilers and pumps can then be operated in a manner that will both maintain supply as well as increase overall efficiency.

Lighting and Doors - Monitoring of lighting’s on/off status as well as interior and exterior door status, open/closed. A better understanding of building occupancy will allow for more efficient decision making as well as overall energy reduction. In addition is the added benefit of an increase in overall security.

Affordable and Scalable

Large initial investments make fast ROI’s difficult if not impossible. Canary offers extreme flexibility in its services and pricing models. Customers can choose from traditional perpetual models as well as lease and subscription services. Canary solutions start as low as $335 a month, and can grow with your operation. Able to collect and monitor as few as 50 data points to millions, the Canary software was designed with every size operation in mind.

Canary and its authorized partners are available to assist with installation and setup, as well as initial and on-going training. Whether your team has years of IT experience or is just beginning to explore building monitoring, Canary has experience that you will find valuable.

Monitor from Anywhere

No matter the device, Canary is designed to deliver information cleanly and without derogating from the user’s experience. The Axiom visualization tool allows users to analyze both real-time and historical time-series process data. Completely customizable, Axiom allows the individual user to quickly create custom dashboards and trend charts, record annotations, build ad hoc calculations, and share their findings with ease. Desktop application will allow for laptop and workstation access, but it is the browser application that offers ultimate flexibility for active users.

Connect to Axiom browser via tablet or smartphone without any need to download and install an application. To increase ease of use, the Axiom browser session is responsive; automatically resizing the screen content to fit your device. To better enhance the experience, the browser session was designed around touch functionality. Now you can easily load charts, add trends, create high/low limits, run ad hoc calculations, and more from your tablet or smartphone.

Ease of Use

According to an active user, “The Canary software is intuitive and requires very little training. Our employees and managers could use the software with very little help. A lot of the managers will use the software to monitor their various building components, looking at screens they designed themselves. These screens are very easy for the operator to look at and see our current power usage and monitor our efficiency. Canary gives us the flexibility to adapt to ever changing conditions, including new technology. We use the system to connect to new devices all the time. There is so much you can actually do. If the Canary system wasn’t there, we wouldn’t have access to a lot of crucial information.”

Read More

Canary Is Becoming Even More Powerful

8:54 AM

New Version Release Set for Summer 2017

Never before has Canary offered so many new tools in a single release. Learn about just a few of the many key developments that will be free to any users currently enrolled in CustomerCare!

Asset Model

Build models defined by custom tag groups and then apply across your entire data base. Quickly organize data in Axiom and move between multiple assets with a single click.


Designed to help you quickly find and store important process events and related data, the new Event feature will allow you to track all critical events and discover how they impact your process. Events will help you track both startups and shutdowns, phases, batches, operator shifts, down-times, and a process that goes beyond acceptable limits. Information captured is specific to the time frame of the event and provides calculated metrics occurring only during the event itself. Use tools like the Excel Add-in or Axiom to explore the data during and around the event; even compare multiple events and report on your findings.

Custom Dashboards

Desire more than basic trending for information displays? Leverage the new, in-application, dashboard editor and create rich Axiom screens on-the-fly. Individual users will love the ability to customize their own layouts and then share their dashboards publicly or save them privately.

Mobile Friendly

Completely redesigned, Axiom can now be easily used on tablets and smartphones. Re-engineered to no longer depend on a "right-click", you can build charts, set limits, and even create ad hoc calculated trends from anywhere.

Excel Add-in Redesigned

Updated and redesigned, the Excel Add-in tool now features Events. Select key assets and view the number of events for a designated time period. Compare multiple events, and gather insight by comparing select processed data that corresponds with the duration of the event.

Help Us Help You!

Canary is seeking specific user requests for feature content and offerings. If you would like to share your thoughts, Canary Engineering would love to hear from you.

Read More

How to Adopt a Predictive Analytic Solution

9:20 AM

SAP recently posted an article titled "10 Myths About Predictive Analytics".  It is a great read and worth your time, and can be found here.  The article makes the following points regarding current Predictive Analytic Myths:
  • 1.  Predictive analytics is easy
  • 2.  Scientific evidence is proof
  • 3.  Only what you can measure matters
  • 4.  Correlation = causation
  • 5.  Predictions are perfect
  • 6.  Predictions are forever
  • 7.  You need a skilled consultant to implement predictive analytics
  • 8.  Predictive analytics is mostly a machine problem
  • 9.  Predictive analytics are expensive
  • 10. Insights = action
A potential client was discussed on a recent phone call with a distributor.  This client manages a fleet of over 100 large sea vessels and is looking for a solution to monitor the entire fleet, as well as move their maintenance model away from interval and run to failure solutions, hoping to employ a predictive or auto prognostic model.
This particular client is looking for an "all-in-one" solution that will help them transition from their current stage, not collecting or sharing any historical process data, to operating under a full machine learning / predictive model.  The challenge is the client is focused on an "out of the box" solution that can offer them everything they want.
You may be chuckling to yourself because you have had interactions with a similar client or you yourself are seeking the same type of solution.  I would suggest a different approach, one that I commonly refer to as the "crawl-walk-run" plan.

The most basic first step is the collection of large amounts of process data.  The top priority of the client needs to be the development of a solution that will allow for all of the sensors that define their many assets to be recorded and stored in a central database.  This database needs to be "loss-less", meaning that the data does not change over time to reduce storage space limitations.  For instance, if you collect a pressure reading at one-second intervals, a "loss-less" database will keep all those individual readings for as long as you like.  Many data historians do not offer this feature.  Instead, after a few months, they turn one second data into 60 second or 5 minute averages.  For some industries or uses, this might be acceptable.  However, if you plan to employ algorithms to scour your data looking for correlations, you don't want to feed it processed time-averages.

Once the data is being stored, the client must put together a plan that will allow for the data to be shared around the organization.  Centralizing data is one thing, but if that data is not easy to access and report from, the information will stay locked in the database with very few individuals gaining any value from it.  To make this possible, the database must be engineered in a way that offers a variety of connections, including ODBC and custom APIs.  It is likely that the process data will be beneficial to combine with other data and move into other systems.

Once the data has been collected and effectively made available to all that can consume it, advanced analytics such as predictive maintenance schedules can be created.  However, to get to this step will require time.  Not because the organization will move slowly, nor because it will be difficult to collect and distribute the process data.  The reason this final step will take time is due to the nature of predictive analytics.  In order for the tools to properly function, it will be necessary to provide years of process data to learn from.  Machine learning is only as good as the data that it is learning from, and having adequate history is a requirement.

Therefore, I made this recommendation to our distributor.  Tell your client to focus on the first two parts of this process.  Today they need to focus less on the analytic solutions that are available, and instead need to focus on finding a strong and reliable database that is capable of storing sub second data from millions of sensors.  That database cannot be SQL based and must be fast.  Recalling the data is going to be of paramount importance for the future process and must be thought about during selection.  Focusing on the best machine learning available today instead of the best data historian available today would be a mistake for several reasons.
  1. Technological advancements - most predictive analytic companies are less than five years old, which is a strong indicator that the market is quickly changing and advancing by the month.  The best solution today is unlikely to be the best solution in two or three years when they are ready to move forward.
  2. Pricing changes - like most new technologies, the price points of predictive analytics have already started to fall, and will further drop over time.  Making decisions today on pricing seems unnecessary when you can nearly guarantee that model will change over the next two years.
  3. Volatile industry - as mentioned in the first point, this is a relatively young market and is full of start-ups.  It seems like an unnecessary risk to form a partnership now when it won't be necessary for another few years.  Given time, leaders will emerge and hold fast, giving a better indicator of who should (and should not) be partnered with.
As with any solution, always start with the basics first.  Find a historian that can handle the millions upon millions of data points you are going to point towards it.  Once the data is collected, then begin to focus on what you can learn and how the organization can benefit.

Read More