The Bad-Order Car Your System Says Is Good

The car has two conditions

A freight car has a physical condition and a recorded condition. The physical car may have a brake defect, damaged safety appliance, wheel condition, overdue test, or another restriction. The recorded car may still appear available to the yard plan, the customer, the interchange partner, or the billing system. When those conditions disagree, the system can keep scheduling work the railroad should not perform.

That disagreement is not a clerical inconvenience. It changes switching plans, train makeup, interchange acceptance, customer commitments, and the work assigned to crews. The operating problem begins when an employee finds the defect. The data-trust problem begins if the finding does not immediately control every downstream decision that depends on the car.

The inspection is only the first handoff

Federal rules establish a disciplined path for moving a defective freight car to repair. Under 49 CFR 215.9, a designated person must determine that the car is safe to move and identify the maximum speed or other restrictions. The train crew must receive written notice. The bad-order tag must identify the car, inspecting railroad, location and date, defect, restrictions, repair destination, and the designated person's signature.

Those requirements show why a simple status field is inadequate. 'Bad order' does not say whether the car is prohibited from movement, safe to move under restrictions, routed to a repair point, repaired but awaiting verification, or returned to service. Each state carries a different operating authority. If another system compresses them into available or unavailable, it discards the information the next person needs.

“The car is not good because the screen is green. It is good when the physical defect is resolved and an accountable person closes the restriction.”

FRA found what the record missed

FRA's January 2025 CSX Safety Culture Assessment provides a useful case study, not an industry-wide defect rate. From January through March 2024, FRA inspected 6,259 freight cars across multiple locations in 22 states and identified 1,946 defective conditions. FRA also reported 572 defects during inspections that included trains that had previously received a Class 1 air brake test. A defective condition count is not the same as a count of defective cars, so it should not be converted into a one-in-five claim.

The larger finding was about the operating environment around the record. FRA observed managers being verbally reprimanded for high bad-order counts, found cars placed into originating trains without required inspections and brake tests, and identified five trains where freight cars were added without receiving the required mechanical inspection and Class 1 brake test. FRA connected the high defect count with inadequate resources and pressure around production. The system did not merely miss conditions. The organization created reasons not to surface them.

A repair record needs a closed loop

A reliable process begins with the physical observation and ends only after an authorized release. The record should preserve who found the condition, where and when it was found, the applicable defect description, photos or measurements when useful, the person who determined movement eligibility, restrictions, repair destination, crew notification, repair performed, and the person who verified return to service.

Every dependent system should consume that controlled state. The yard plan should not assign a restricted car to ordinary work. Customer visibility should not promise availability the railroad cannot support. Interchange records should carry the restriction instead of forcing the receiving railroad to rediscover it. Corrections should preserve history, because an unexplained status change cannot prove that the physical defect was repaired.

Measure whether the exception stays visible

Start with one terminal or industrial switching operation. Follow every bad-order car from detection through disposition. Count how often crews find a condition not reflected in the system, how long restrictions take to reach the work plan, how often cars are reassigned after a defect is entered, how long repairs wait, and how often release records lack evidence or accountable approval.

The goal is not to reduce the bad-order count. A falling count can mean better equipment, or it can mean the operation stopped reporting what it found. Measure whether defects are discovered, communicated, protected, repaired, and closed without being lost between departments or railroads. The good system is not the one that shows fewer exceptions. It is the one that keeps the real exception in control until the car is truly ready.