Explore solutions built for your industry

Our customer-proven solutions monitor medications and food inventories for some of the most recognizable names in the industries of healthcare, food service, and transportation, and logistics. See how our solutions adapt to your industry needs.




Explore solutions for pharmacies, and laboratories.

See solutions

Food Service

Explore solutions for restaurants, grocery, and hospitality.

See solutions

Transportation & Logistics

Explore solutions for meeting FSMA/CDC compliance.

See solutions

System Overview

Share SmartSense Solutions with your team.



ROI Calculators

See how much time and money you can save with SmartSense.

Calculate ROI


Review technical specifications for our solutions.

See datasheets

Resource Center

Work smarter. Explore our videos, webinars, and customer stories.

See resources

Questions? Contact us.

Call +1 (866) 806-2653 to speak with our experts or get started with a demo.


About Us

SmartSense was created to use the power of the Internet of Things (IoT) to help our customers protect the assets most critical to the success of their business.

See our story


Create the future of IoT by joining our team.

See job openings


July 26, 2021

Tagging and Labeling Assets Unlocks Additional Value from Your Data

Written by Jason Sroka | Data

One of the best practices we highly recommend to every new customer we work with is to take time and care to label all of their cooler and freezer units—also known as “assets”—during installation. There are big benefits to be gained when asset labeling is done in a planned and systematic way because it allows you to tackle strategic and long-term questions with a little investment in design and process upfront. While adding labels to an existing deployment is readily possible, first installation is an ideal opportunity as you’ll be visiting each asset to install sensors and can capture detailed information like make and model, asset contents, and many other things.

Labeling assets sets the stage for your data to produce powerful insights and drive smart business decisions. For instance, three frequently used, cross-vertical labels are "environment," "make," and "model." Do you want to be able to assess whether your coolers or freezers generate more alerts? Use an Environment tag with those labels and you’ll have the ability to answer that question. Want to know how much more frequently open-air coolers go out of range relative to reach-ins (and walk-ins, and bunkers, and service cases…)? Again, set up a tag for equipment type, use those labels to annotate your equipment, and you’ll be able to answer that question. Want to know which Make and Model of fridge experiences the fewest issues over the long-term so you are more informed the next time you need to purchase equipment? Use Make and Model tags and you’re on the path to an improved procurement approach!


Leveraging tags and labels make it easy to understand head-to-head equipment performance.

Tagging and Labeling

Labeling is actually a two-step process. The first step, Tagging, establishes the attributes or “Tags” you’re going to capture. Examples of tags from above include the type of equipment it is (e.g. open-air cooler, walk-in fridge), the manufacturer and model of the equipment, the target environment (e.g. cooler, freezer, prep room, ambient), but anything else that could prove useful when reviewing data from the equipment is fair game. When you think about the different ways you might want to break out your data for comparisons across subsets of your assets, each of those ways represents a potential Tag (though some tags can support multiple prospective analyses, and in some cases like Make and Model it can be useful to use multiple tags to support a single type of analysis).




Labeling then describes the process of assigning values to the tags. Often (though not always) this involves deciding ahead of time what values will be allowed. Most of our customers use Environment tags as part of configuring their alerts, with a mix of Coolers and Freezers, and often also Ambient air conditions, but also need to cover less-common situations like Prep Rooms in groceries or cryo units in healthcare.  Make and Model have become standard Tags to the extent that we’ve made it the default that customers will have access to two tags for Make and Model. The SmartSense platform allows you to add new Tags as you recognize additional useful features to break out your assets by. 


Asset editing screen in SmartSense
SmartSense makes it simple to capture tags and labels during the setup process.


An ideal time to tag and label is upfront during the installation of sensors in each of your assets. Each piece of equipment needs to be visited in order to install the sensor(s), and it’s a great opportunity to systematically go through a set of tags for each piece of equipment and introduce a standard approach to labeling them. But that’s not the end of the journey, because all of our clients eventually have new equipment or replacement equipment and to get the full value from tags and labels you need to ensure that the data will be accurate long after the initial install.


Designing a Process

Tagging and Labeling should begin with an analysis of the questions that you want to be able to answer using your data, as discussed above. But even when you’ve decided on those goals and have a sense of the set of tags you want to introduce, there are still big questions to answer about the labels you will allow.


By predefining a set of labels, you avoid a lot of potential issues around data collection, such as typos and misspellings that cause problems when you then attempt to roll up by label.


In some cases, a complete set of labels are defined and there is no opportunity to extend beyond them. This is perfectly reasonable in situations where the set of options are known at the start and won’t change, for example, if you know that you only have Cooler, Freezer, and Ambient environments. At the other extreme, we have worked with clients who defined Tags for the Make and Model of their equipment and allowed free text entry of labels with no reconciliation against a predefined list. 


By predefining a set of labels, you avoid a lot of potential issues around data collection, such as typos and misspellings that cause problems when you then attempt to roll up by label. When labelers have free text options, there is often a lot of cleanup required to get all the equipment that should fall into a single category to actually align into that category, especially for Model numbers that include dashes that labelers might consider optional. A predefined list of options avoids those issues. However, you don't always know all the right labels at the start, so while you may want to avoid the challenges of not predefining labels you may also need to have a way of extending the list—for example, adding another equipment provider if you purchase a new type of equipment.


Extensibility can be achieved by periodically reviewing labeling options for tags and determining which ones should be extended. This works in a case where the labels needed can be known beforehand, for example, if the procurement team announces there’s a new manufacturer they are purchasing equipment from and has the label set extended to reflect the new option. In a lot of cases, this foreknowledge is not practical.

We have worked with clients that have thousands of sites and a menagerie of equipment in them that has accumulated over time and that no one would be able to generate a comprehensive list for in advance. In those cases, you will want to design a process for allowing the people doing the labeling to suggest extensions. As an example here, we have implemented approaches where a predetermined set of labels are provided as options and then the labeler has an “Other” option which, if selected, allows them to submit free text for the label. This label then gets flagged as a potential new standard value and a periodic review process checks each of these new potential labels and determines if the labeler should have mapped to an existing label (this tends to get increasingly common over time) or if the new label will be accepted and promoted to a standard label option.


What this highlights is that the best practice of tagging and labeling does not end after the initial rollout, but must be implemented in a manner that is robust to new asset replacement and purchase and to the evolution of your business operations.


What this highlights is that the best practice of tagging and labeling does not end after the initial rollout, but must be implemented in a manner that is robust to new asset replacement and purchase and to the evolution of your business operations. It also means going into your data sets and changing or adding tags and labels when you have new business goals to achieve, e.g. what once was a single category (e.g. ‘Cooler’) might need to be split into multiple categories (‘Cooler – Meat’ and ‘Cooler – Produce’ reflecting that although the safety ranges may be the same the target temperatures and relative risks of events among the two may warrant separate categories). 


Similar to how you want to design a process for label options to be updated, you want to ensure you have a process for maintaining the labels’ accuracy. You can't just take a snapshot now and then assume it's going to be accurate years later, especially when sensors from an old fridge are put in a new replacement fridge that could be a different brand of equipment. To keep your analysis as accurate as possible, you need to create and adhere to a robust standard operating procedure (SOP) that not only inputs accurate and detailed tags and labels at that first rollout but also keeps the tags and labels up to date. With every new fridge added to your fleet or fridge replacement, you want to ensure that they will be correctly labeled across all the necessary tags. 


Consider Data Goals When Selecting an IoT Solution 

When you're choosing an IoT solution, you want an organization that can truly partner with you to identify and anticipate your labeling requirements at the onboarding and installation phase. SmartSense has the expertise and flexibility to guide you through a proper strategy and the creation of an SOP.

We know what's important to your vertical, to your customers, and to your executive management. We can suggest solutions you may not have thought of yourself. Because we have years of experience working in these verticals, we also have comprehensive knowledge of what labeling categories and tags already exist and would work best to fulfill your business plan. As an example, we can generate lists of Makes and Models that other companies in your industry have used as a starting point when deciding what labels you will initially allow.

To empower you to manage the process independently, we’ve programmed that expertise directly into the SmartSense Install App. It’s a terrific tool that lets you choose or create the best labels and tags that will generate the right data to help you make the right decisions to succeed in your specific business, market, or industry.

Thinking ahead to the kinds of questions you want your data to be able to answer helps guide you to establishing a robust set of tags and labels that will support it, and designing a robust process for implementing and maintaining the labels will ensure that you can have confidence with what the data then tells you. Reach out to us if you have any questions or want to discuss best practices we’ve seen from similar customers in the past – we love to help our customers get all the value possible out of the data we’re collecting for them!


Jason HeadshotJason Sroka is the Chief Analytics Officer at SmartSense by Digi. He earned B.S. degrees in managerial science and electrical engineering at Harvard, and a Ph.D. in Speech and Hearing Sciences at MIT. If you would like to contact Jason with any questions about this post, please email him at Jason.Sroka@digi.com.


Topics: Data

Subscribe to the SmartSense Blog

Stay up-to-date with the latest news in food and pharmacy safety, facilities monitoring, and supply chain visibility.