Figure 1. The GM V8 assembly process at the Romulus Powertrain Engine Facility terminates with an extensive, 15-mile-long conveyor system that automatically sorts finished engines for shipment. Dan Sinclair is shown using a FLIR ThermaCAM® infrared camera to check the condition of the rollers and the chain.

Nearly all electro-mechanical equipment becomes anomalously warm before it fails, making infrared (IR) cameras extremely effective diagnostic tools in the manufacturing environment. Inspections using infrared cameras can find many problems before failure occurs. In many cases, the time to failure can be projected, enabling the most convenient scheduling of proactive or preemptive repairs. This practice, called “predictive maintenance” (PdM), enhances both productivity and safety.

IR cameras play a major role in PdM programs in manufacturing plants, electric power transmission and distribution systems, chemical plants, paper mills, and numerous other industrial operations. IR cameras are also ideal for monitoring objects and materials that present diagnostic thermal profiles, such as electricity transmission and distribution systems, material in containment vessels and pipelines, materials and associated equipment during the manufacturing process, and breaches in security. Other well-regarded inspection tools include human senses, vibration analysis, oil analysis (tribology), and ultrasound analysis.

Figure 2. During a walk-through inspection, an anomaly was spotted thermally (left) and visually (right) in a roller in this turn.
Figure 3. Upon closer inspection, the infrared camera pinpoints the actual failed roller, which has an anomalously warm temperature of 103.9 °F. Photo at right shows the roller close-up.

IR thermal inspections are accurate, rational, intuitively interpretable, non-destructive, noninvasive, noncontact, and fast. They provide instant images and data that are immediately usable in reports, and they can be easily archived to maintain a trending study of performance, which in turn may be used to project time-to-failure, enabling optimal scheduling of maintenance, based on actual operating condition, and pre-empting catastrophic failure.

Figure 4.
Figure 5.
Figure 6.
Figure 7.
Figure 8.

Infrared technology has been used at the General Motors Powertrain Engine Facility in Romulus, MI on a full-time basis since 1988 as part of GM’s predictive maintenance program. Infrared inspections add real value to the total predictive/preventive maintenance program. Electrical components — including the aging electrical buss, all mechanical equipment, and the building envelope, including the roof — are continually inspected.

Through a corporate initiative, the GM Infrared Standards Committee continually tracks the value of this program on the basis of a written cost-avoidance calculation and procedure. As a result of continued, demonstrated savings, GM has adopted a written practice that is treated as a living document. Following are three recent cases at the Romulus facility in which inspection of equipment using FLIR infrared cameras has yielded significant savings.

Engine Delivery System Turns

The V8 engine assembly process is terminated on a “power and free system” in which the finished engines are marshaled and sorted for shipment by automated equipment. Within the system, many dips and turns have been incorporated to facilitate the 15 miles of chain needed to accomplish this task. Turn roller failure (Figures 2 and 3) and an overheated chain (Figures 4 through 7) were two problems that were resolved with proactive repairs before they could cause major downtime. If allowed to run to failure, these problems would have necessitated more costly and time-consuming reactive repairs.

Overheated Chain

While examining the west chain on the V8 engine track (Figures 4 and 5), it was noticed that it was about 10 °F warmer than the east chain (Figures 6 and 7). An elevated temperature indicates friction, which causes wear and increased electrical load. After a short investigation, the culprit was discovered — an empty automatic grease system, which was promptly refreshed.

Overheated Bearing

While examining the west chain on the V8 engine track (Figures 4 and 5), it was noticed that it was about 10 °F warmer than the east chain (Figures 6 and 7). An elevated temperature indicates friction, which causes wear and increased electrical load. After a short investigation, the culprit was discovered — an empty automatic grease system, which was promptly refreshed.

Proactive Versus Reactive

The success of any preventive maintenance program is measured by comparing the costs that are avoided by early, proactive detection and optimal scheduling of repairs vs. the costs of making reactive repairs after a failure or breakdown occurs. Proactive repairs make the same common sense as locking the proverbial barn door before, rather than after, the horses are gone. At the Romulus engine plant, there is a formal Cost Avoidance Worksheet for proactive repairs, which includes:

  1. A statement of the “root cause for action” — a description of the imminent problem
  2. A description of the proactive repair performed
  3. An analysis of the costs of the actual proactive repair task
  4. An analysis of the projected costs of the reactive task that was preempted
Figure 9.

By comparing the costs of the proactive task with the projected costs of the reactive task, the value of the preemptive repair is determined. While some cost factors may remain unchanged in both cases, e.g., replacement of worn-out and broken parts, three major benefits can be realized by proactivity. First, repairs can be scheduled during convenient times, such as during shift changes or during planned downtime.

Second, the collateral effects of actual failure are avoided, such as additional damages, production losses, and worker safety issues. Third, the time required to make proactive repairs is likely to be substantially shorter than reactive repairs, further minimizing or eliminating lost production. With this in mind, here are analyses of the costs avoided in two actual examples of proactive repair.

Overheated Chain – Case 2

The cost of proactive repairs to the overheated chain included $45 for one man-hour of labor plus $20 for grease. Compare the “projected reactive task costs” that would have accrued had the chain been run to failure: 136 hours of repair labor at $45 per hour, plus 1,072 hours of lost production labor at $39 per hour.

Add parts costs of $32,430 to replace 4,600 links of chain, $750 for new drive chains, and $20 for grease. The avoided costs were clearly substantial. The total cost of the proactive repair was a mere $65, but the total reactive repair would have cost $81,078, plus the unquantified “cost” of lost production for two shifts — an estimated 2,100 V8 engine units — for a total savings in the range of $1 million.

Anomalously Hot Feeder Buss Isolation Switch

Figure 10.
Figure 11.

Infrared inspection revealed that a feeder buss isolation switch on the north side of the plant was emitting a heat signature consistent with a high-resistance connection. The switch was shut down, locking out the buss supply. The feeder buss was disconnected from the switch, the buss bars were cleaned, buss insulators were replaced, and buss power was reset without incident. The cost of this proactive repair was $1,080 for 24 man-hours of repair labor at $45 per hour, and $590 for replacement buss insulators — a total of only $1,670. If the situation were ignored and run until failure, the reactive repair costs would have totaled $41,977, the sum of 156 hours of repair labor at $45 per hour, 672 hours of lost production labor at $39 per hour, and $8,749 for parts, including replacement buss insulators, a section of buss, a new buss isolation switch, temporary service wire, a 1000A temporary buss plug, and a 1000A fuse.

The savings totaled $40,307, plus the avoided costs of ripple effects throughout the section of plant crippled by the resulting unplanned power outage. These avoided costs would include the cost to recover machine programs, cost of cutters and tooling destroyed due to in-cycle power failure, and the significant but incalculable cost of in-process engine blocks that would have to be scrapped due to tooling failure.


An ounce of PdM prevention is worth a ton of cure. Dramatic savings are achievable from regular preventive and predictive maintenance in large manufacturing facilities, based on detection of incipient failures by infrared thermal inspections and other test methods. The savings accrue from three perspectives:

  • First, repairs are much cheaper to make before catastrophic failure occurs, and such proactive repairs avoid collateral damage to other equipment, in-process product, and even personnel.
  • Second, repairs can often be made during scheduled downtime or during shift changes, minimizing or eliminating lost production.
  • Third, the time required to make proactive repairs is likely to be substantially shorter than to make reactive repairs, further minimizing or eliminating lost production.

The dollar value of proactive savings has traditionally been difficult to certify, but modern “smart” infrared cameras and other computer friendly test equipment are greatly facilitating record-keeping by downloading data to easily analyzed digital archives. In addition, report-generating and data-archiving software such as ThermaCAM Reporter and DataBase from FLIR Systems are greatly facilitating the quantification of proactive repairs at General Motors. The result is an ongoing paradigm shift by enlightened plant management. The plant maintenance paradigm is moving from the cost side of the ledger, where it has traditionally been considered “overhead” Maintenance & Operation (M&O) costs, to the nascent category of Avoided Costs.

This shift is in turn recasting maintenance professionals and PdM programs as part of the profit-making side of today’s manufacturing organizations. Indeed, a dollar saved has always had the same value as a dollar earned. In a highly competitive global industry such as automobile manufacturing, the true value of Avoided Costs produced by today’s predictive/preventive maintenance professionals is realized on the income side of the ledger, and can be expressed in terms of dollars-worth of increased productivity, lower manufacturing costs, and larger margins.

This article was written by Daniel Sinclair of General Motors Corporation (Detroit, MI) and Leonard Phillips of FLIR Systems (Boston, MA). For more information, Click Here .