Best practices for taming data
The goal should be to turn data into information, and the best way to do that is via data integration.
There are so much data available to dairy processors that it can feel overwhelming when trying to determine what needs to be tracked and what to ignore. The reality is that as we move into the future, even more data will need to be evaluated and managed. In fact, analyticstraining.com says that by 2020, data are expected to have increased some 430%.
One big area where data are critical is the transportation portion of a business, which is often not the business’s core competency. This can be especially important when dealing with products that are temperature sensitive and have freshness windows.
The data problem is exacerbated by the fact that there are myriad sources that provide data on vehicle health and location in different formats, from DVIRs, telematic devices and electronic logging devices to maintenance records, OEM portals, third-party service providers and breakdown service companies. The list of data sources may seem endless. In fact, there are so much data, most companies barely tap into a fraction of what they collect.
Turning data into information
Data from disparate sources and in different formats are not necessarily useful. The goal should be to turn data into information, and the best way to do that is via data integration. When you have information rather than simply data, you can leverage that information to make informed business decisions.
Getting data under control as they pertain to the transportation function starts with having a data-management process in place. You need to determine what data are going to be most beneficial to your operation. Here are some categories you should consider.
- Operational information: types of loads, weights of loads, number of stops, etc.
- Routing information: terrain, distance traveled, weather conditions
- Maintenance information: scheduled preventive maintenance and maintenance history
- Telematics: equipment use, fault codes, driving conditions, third-party repair shop availability, repair status, etc.
You also need to begin making decisions on the granularity of the data you will need. This includes things like year/make/model of equipment, route/driver identification, regional differences and in-house vs. outsourced maintenance.
Data collection process
Once you know which data you want to collect, you need to develop a process for analyzing that data. Start by identifying everyone who will be involved in the process and all key stakeholders. Consistency and accuracy of data are critical, so try to avoid manual data entry. When data are entered manually, there is more likelihood for errors to occur.
The data collection process needs to establish a history and include ongoing monitoring of process performance. Collected data should include key performance indicators (KPIs) and supporting data with time stamps from multiple points within the process lifecycle at consistent intervals and/or process transfer points.
You will need to develop key performance indicators (KPIs) to spotlight weak points in lines of business that fail to meet expectations, regulations, production quotas or efficiency standards. KPIs can provide triggers for first points of contact to analyze and diagnose incidents. Balance issue coverage with highlighting areas meeting expectations, and determine who could mentor underperforming ones. Some common KPIs in the dairy-processing world include the following:
- Tractor and trailer utilization: This tells you how full trucks are going out every day. In other words, pieces or gallons per load.
- Tractor miles per gallon.
- On-time performance: This is a customer-based metric and measures whether you are delivering product within the established delivery window.
- Average stop time and delay time: Where in the delivery network are there delays? Understanding where the delays are occurring and what is causing them can help mitigate or eliminate them.
- Temperature tracking: This looks at trailer temperature but also product temperature.
Adopt a data governance strategy to collect, transform and store data efficiently and securely. Ask users of the data about the level of detail they desire and their preferred delivery method.
One common approach is to provide a dashboard website with high-level statistics about critical KPIs and set up alerts to send notifications out to first point of contact when thresholds are exceeded. Setting up dashboards with red, yellow and green indicators allows users to spot critical issues quickly and work to resolve the most critical ones before moving on to less important ones.
Providing users with a method for drilling down into the details of each KPI alert should enable users to determine the reason for the alert or provide more information. One word of caution: KPIs must focus on things that are actionable so that you can adjust your operations to make improvements as necessary.
Prompt, relevant and actionable data enable you to take action before an incident becomes a liability. The right data can help you make decisions regarding future equipment purchases, driver and technician performance and training needs, asset replacement cycles and more.
When it comes to data, it is important to remember that the output is only as good as the input. Details matter; the more granularity, the better. It is not just a matter of turning on data flow, but rather having the right data and the right people in place to analyze the data to improve operational efficiencies.
