Are your APCs ‘down’ for the count?
Maximize the efficiency of your automatic passenger counters for NTD certification-worthy results.
By Matt Schroeder
Director of Marketing, ETA Transit Systems
I am going to dispel a myth for your right off the bat.
The FTA (Federal Transit Association) does not provide certifications for APC hardware. Never have, never will. Anyone that tells you that their brand of APCs has been certified is straight-up lying.
It’s amazing to us just how many transit operators are under the misconception that there are specific brands of APC hardware blessed with a governmental seal of approval.
APC suppliers may provide you with a process for certifying the counts generated by your automatic passenger counters, but that’s it. Certification of counts is a whole different beast.
In general, the certification process entails two things:
- An APC benchmarking plan for the first year
- An APC maintenance plan for every three years
You can read all about it here: NTD Policy Reporting Manual
What defines a benchmarking plan?
Benchmarking requires validation for unlinked passenger trips (UPT), and passenger miles traveled (PMT) data against a manual sample to prove the APC system’s accuracy. The process for reconciling this data includes:
- For agencies with fewer than 30 vehicles, the sample size must include at least 15 trips on APC-equipped vehicles.
- For agencies with 30 or more vehicles, the criteria changes to the larger of 15 trips or half the number of APC-equipped vehicles (up to 50 trips).
- In both instances, the trips do not need to be randomly selected and may spread over a year. We recommend that these samples include times of heavy ridership and the inclusion of at least one trip per vehicle type and APC model.
- A description of the agency’s APC systems and sampling procedures is also required.
- Should a trip be excluded or rejected from a provided sample, an explanation is required.
- Also to be included is documentation that accounts for the percentage of trips made that do not have valid APC data due to equipment malfunction, data corruption, failed verification checks, or any other reason.
For manual counts that use data collection staff or onboard cameras, the FTA strongly recommends a data collector at each door during heavily loaded trips. It is appropriate to process APC data to correct for anomalies, just as it is reasonable to adjust for reporter data during the collection process. The goal is to compare manual and automatic counts to demonstrate equivalency or that any differences are justifiable.
PMT calculations should utilize the same set of distances for all analyses. Should any differences exist (i.e., interstop), an explanation must be included with the provided data.
Acceptable deviations in data are less than five percent for:
- Difference between manual vs. APC UPT counts
- Difference between manual vs. APC PMT counts
If a transit agency uses APCs for both directly operated and purchased transportation services, a requirement for separate samples of each type of service exists.
Automatic counting hardware is exceptionally precise, with most manufacturers boasting 95 to 98 percent accuracy. Despite this precision, errors can creep into the system, which is why the FTA does not certify the hardware itself. Simply put, nothing is going to deliver optimal accuracy out of the box. Configuration, tweaking, and adjustments are typical to account for environmental variables unique to every transit operation.
The process of verifying your counts is quite simple.
Option 1: Manual counts
Your agency can have an employee ride every route and take manual counts to compare to those generated by the APC. This process quite literally involves a person with a clipboard taking note of who boards and alights at every stop, the pencil and paper method.
Option 2:Digital counts
If a vehicle has a mobile data terminal (MDT), the driver can use a built-in digital passenger counter to log passenger boardings and alightings. The data automatically uploads to the intelligent transit system for future review.
Of the two, we recommend the MDT-based approach. It’s less prone to data entry errors, and the information is easily reviewable in any reporting or analytics software for side-by-side comparison against the APC-generated counts.
Once the necessary amount of historical count data is collected, you will work with your transit technology provider to develop a reporting algorithm that accounts for recurring variations in the information. This final step refines the configuration of your APC and subsequent analysis and allows for the consistent benchmarking required by the FTA for certification. It also serves as a diagnostic resource to evaluate the consistent performance of your hardware; a sharp increase or drop-off in counts could be an indication of a hardware error or mounting issue.
Adding automatic passenger counters is a great way to reduce manual processes and gather the critical data you need to understand passenger travel behavior. The process is not straightforward and requires a few essential steps to ensure that the counts you collect provide your agency with the reliable counts you need to make informed decisions and submit accurate reports to the National Transportation Database (NTD). It is essential to understand that the hardware is not certified by the FTA but rather the methodology for collecting, analyzing, and reporting the data that receive certification. This process takes time, but when working with a reliable transit technology provider, your new APCs will provide trustworthy information you can use to make informed operational decisions for years to come.