OPINION
BAKER, Judge.Appellant-petitioner Union Tank Car, Fleet Operations (UTC) appeals the trial court's affirmance of the Board of Safety Review's (Board's) assessment of five knowing violations of the Indiana Occupational Safety and Health Act of 1972 against UTC. UTC contends: 1) the Board applied an incorrect definition of "knowing" and 2) the Board's assessment of knowing violations is contrary to proper procedures and unsupported by substantial evidence.
FACTS
The facts most favorable to the Board's decision reveal that UTC is engaged in the manufacture, leasing and maintenance of railroad tank cars. To maintain its cars throughout the United States, UTC operates mobile units which are based at various host facilities and serve an area within a 100 mile radius of the host facility. During 1989, UTC operated a mobile unit at the Amoco Oil Company refinery in Whiting, Indiana (Whiting Mobile Unit). Four UTC employees were assigned to this unit, including the unit's supervisor, James Greuner, and three tank car repairmen, Danny Bouchee, Douglas Anderson and John Carter. However, because of an illness, Greuner was not physically present to supervise the Whiting Mobile Unit during the Summer of 1989; rather, he attempted to supervise the employees from his home or the hospital via the telephone. Although UTC was aware that Greuner was not actively supervising the unit, it did not temporarily or permanently replace him.
The duties of a mobile unit include performing exterior repairs and interior "wipe downs" on the tank cars. Wipe downs involve removing excess debris from inside the empty tank cars. Prior to entering the tank cars to perform wipe downs, mobile unit employees are required to test the interiors of the cars to ensure that the car atmospheres are non-toxic and are not oxygen deficient. To promote testing of the tank cars, UTC developed a Confined Space Entry Program for its employees. Specifically, the program requires employees to complete a tank entry permit prior to entering the railcar by testing for "oxygen, flammability and-if required by the Tank Car Cleaning Instruction Manual, or the Safety Department-toxicity." Record at 205. The employees are then required to post the permit outside the car. To test the atmosphere of a railcar, employees use an Eeolyzer, which is a device used to monitor combustible gas and oxygen deficiency. Despite the program developed by UTC, however, the evidence revealed that Greuner instructed employees of the Whiting Mobile Unit to disregard the testing procedures proscribed in the tag entry permit system.
On August 1, 1989, an Amoco employee attempted to prepare a tank car for loading when he noticed pressure inside the car. After loosening the top manhole cover, the employee visually inspected the car and noticed debris in the bottom. He notified the Whiting Mobile Unit that the car needed to be wiped down. Thereafter, Carter and Anderson performed a wipe down of the tank car. When they did not return to the UTC office, Bouchee went to the worksite to look for them and discovered Carter and Anderson lying inside the car. He immediately informed the Amoco employees, who rescued the men from the tank car. Anderson was hospitalized for ten days and *888suffered permanent memory loss due to oxygen deficiency. Carter died from lack of oxygen.
On January 24, 1990, the Commissioner of Labor filed a Safety Order and Notification of Penalty (Order) against UTC pursuant to Indiana's Occupational Safety and Health Act (IOSHA).1 Specifically, the Order alleged ten "knowing" violations of IOSHA's general duty clause by UTC. UTC challenged the Order and a hearing was held in front of Administrative Law Judge Barry Macey (ALJ). On September 29, 1992, the ALJ issued his recommended decision, upholding five of the Order's knowing violations. These violations included:
(1) UTC violated the general duty clause because it did not identify and evaluate each potential hazard of the permit space, including severity.
(2) UTC violated the general duty clause because it did not provide, maintain, and ensure the proper use of equipment necessary for safe entry including testing, monitoring, communication and personal protective equipment.
(3) UTC violated the general duty clause because it did not ensure that employees received training such as () knowing the hazards which may be faced during entry; (ii) recognizing the signs and symptoms of exposure to a hazard; (Hi) understanding the consequences of exposure to a hazard.
(4) UTC violated the general duty clause because it did not establish and enforce a written permit system for the proper preparation, issuance and implementation of entry permits.
(5) UTC failed to explain the hazards of nitrogen and white oil and did not explain specific procedures the railcar repairmen could take to protect themselves from exposure to nitrogen and white oil.
R. at 1129-67. As a result of these violations, the ALJ imposed penalties totaling $42,000.00. Thereafter, the Board adopted the ALJ's findings. UTC appealed the Board's decision to the Marion Superior Court, which, on April 18, 1995, affirmed the Board. UTC now appeals.
DISCUSSION AND DECISION
I. Definition of Knowing
First, UTC contends that the Board applied an incorrect definition of "knowing" in assessing violations against UTC. Specifically, UTC argues that the definition of knowing used by the Board, that of plain indifference, fails to contain the element of a "bad motive" on the part of an employer. As a result, UTC claims that the definition of a knowing violation is indistinguishable from that of a serious violation, which carries with it a much lesser sanction. Thus, UTC urges us to overrule the definition of knowing affirmed by this court in Commissioner of Labor v. Gary Steel Products Corporation, 643 N.E.2d 407 (Ind.Ct.App.1994), which does not contain the heightened motive requirement.
During 1989, IOSHA provided for different levels of violations, including nonserious, serious and knowing. However, IOSHA failed to include definitions or standards by which each level of violation should be proved. See IND.CODE § 22-8-1.1-27.1. As a result, the Board in Commissioner of Labor v. Auburn Foundry, Inc., No. 40C01-9008-MI-2958 (July 19, 1990), defined a "knowing" violation to include an "obstinate refusal to comply" or a "flaunting of the act." Then, in Gary Steel, the Board used a definition of knowing which only required a showing of plain indifference or intentional disregard of the welfare of employees to prove a knowing violation, which this court affirmed. 643 N.E.2d at 412. In doing so, this court cited with approval Georgia Electric Co. v. Marshall, 595 F.2d 309 (5th Cir.1979), in which the Fifth Cireuit Court of Appeals found that a bad purpose requirement was unnecessary to distinguish between serious and knowing violations. Gary Steel, 643 N.E.2d at 412. Specifically, the Fifth Circuit found that while a knowing violation required a showing that the employer acted voluntarily, a serious violation only required a showing that there was a substantial probability that a particular violation could result in death or serious *889physical harm. Electric, 595 F.2d at 318-19.
We agree that the definition affirmed in Gary Steel, that of plain indifference, is adequate to show a knowing violation by an employer. Further, we agree that the definition permits a distinction between serious and knowing violations. We decline, therefore, to revisit this issue.
II. Knowing Violations
Next, UTC asserts that the Board's finding of five knowing violations is contrary to proper procedures and not supported by the evidence. We note that in the instant case we are reviewing the decision of an administrative agency. IND.CODE §§ 4 21.5-5-1 to 4-21.5-5-15 of the Indiana Administrative Adjudication Act establishes the exclusive means for judicial review of an agency action. A court reviewing an administrative decision is limited to determining whether the ageney possessed jurisdiction over the subject matter and whether the agency's decision was made pursuant to proper procedures, was based upon substantial evidence, was not arbitrary or capricious, and was not in violation of any constitutional, statutory or legal principles. Indiana Dep't of Natural Resources v. United Refuse Co., 615 N.E.2d 100, 103 (Ind.1993). The trial court proceeding is not intended to be a trial de novo, but rather the court simply analyzes the record as a whole to determine whether the administrative findings are supported by substantial evidence. Id. Courts that review administrative determinations, at both the trial and appellate level, are prohibited from reweighing the evidence or judging the credibility of witnesses and must accept the facts as found by the administrative body. Peabody Coal Co. v. Indiana Dep't of Natural Resources, 629 N.E.2d 925, 928 (Ind.Ct.App.1994). However, we need not accord the same degree of deference to an agency's conclusions on a question of law. Id.
A. Retroactive Application of Standards
First, UTC contends that for each of the violations found by the Board, the Board formulated a standard and procedures for a confined space entry program and then retroactively applied them to UTC. According ly, UTC asserts that the Board violated its right to due process and erroneously engaged in administrative rulemaking rather than adjudication.
To satisfy due process, administrative decision making must be done in accordance with previously stated, ascertainable standards. Dept. of Environmental Management v. AMAX, 529 N.E.2d 1209, 1212-13 (Ind.Ct.App.1988). The standards should be written with sufficient precision in order to give fair warning as to what the agency will consider in making its decision. Id. at 1218.
In finding violations by UTC, the Board set forth examples of procedures or programs which were absent from UTC's safety program. For example, the Board noted that UTC's program did not contain an independent requirement that its employees test the atmospheres of confined spaces; rather, UTC's testing requirements were contained within its tag permit entry system. R. at 1135-86. While this finding appears to contain a standard against which UTC's program was compared, as the State notes in its brief and during oral argument,2 the Board's findings were proffered as acceptable procedures and programs, not as requirements. UTC was not assessed violations because its programs failed to contain certain elements; instead, the Board found that UTC's confined space entry program was deficient on its face and improperly implemented. The Board's subsequent adoption of standards for a confined space entry program had no effect on UTC, as UTC's program was not compared to these standards to find violations. As a result, we reject UTC's argument that the Board engaged in retroactive rulemaking or that it imposed standards on UTC which were not ascertainable.
B. Sufficiency of Evidence
Next, UTC challenges the sufficiency of the evidence which supports the Board's determination that UTC knowingly *890violated its general duty to protect its employees. The general duty clause provides:
Each employer shall establish and maintain conditions of work which are reasonably safe and healthful for employees, and free from recognized hazards that are causing or are likely to cause death or serious physical harm to employees.
IND.CODE § 22-8-1.1-2. To prove a knowing violation, the Commissioner of Labor is required to prove that UTC acted voluntarily, with either an intentional disregard of, or plain indifference to, its employees. Gary Steel, 648 N.E.2d at 411. According to UTC, while the evidence could support a "serious" violation, it does not show an intentional disregard or plain indifference by UTC to the welfare of its employees.
UTC contends that its numerous programs and procedures for regulating entry into confined spaces demonstrates that it made a good faith effort to protect its employees. As a result, UTC argues that it did not violate its general duty to: 1) identify and evaluate potential hazards, 2) provide and maintain equipment for safe entry, 3) train employees regarding hazards, and 4) implement a written permit system. Specifically, UTC asserts that during the hearing before the ALJ, it presented evidence that it had developed a confined space entry program which required employees to test the atmosphere of tank cars prior to entering them and to attach a tag permit on the exterior of the cars noting that the testing had been completed. Although the confined space entry program did not require employees to communicate with each other while working on tank cars, the employees normally utilized a system whereby one employee would remain outside the tank car at all times to maintain contact with the employee working inside the car. Additionally, UTC presented evidence that it developed a program to train its employees to use Ecolyzers to test the atmospheric oxygen levels inside rail cars.
UTC also challenges the Board's determination that it knowingly failed to explain the hazards associated with nitrogen and white oil. At the hearing, UTC presented evidence that it required each facility, including the mobile units, to maintain material safety data sheets (MSDS) for all hazardous chemicals used in operations.3 Further, UTC contends it trained each employee regarding the use of hazardous chemicals. In fact, at the Whiting Mobile Unit, the employees had signed a certificate indicating that they had received this training. In addition, UTC presented evidence that it had an MSDS for white oil, although it is not a hazardous chemical, and an MSDS for nitrogen, although it was not commonly used to off-load tank cars.
Notwithstanding UTC's evidence and argument to the contrary, the remaining evidence presented at the hearing supports the Board's determination that UTC was aware of the hazards of its employees entering confined spaces, particularly those spaces which had previously contained potentially toxic chemicals, yet made only illusory or superficial efforts to deal with those hazards. Specifically, the evidence revealed that at the Whiting Mobile Unit, UTC's procedures and programs were ignored. UTC's investigation after the accident revealed that Greuner, the supervisor at the Whiting Mobile Unit, instructed his employees that it was unnecessary for them to complete the tag permits prior to entering the tank cars. R. at 4264, 4220. Additionally, it is apparent that neither Carter nor Anderson actually tested the air inside the tank car in which the accident occurred prior to entering it. Had they done so, the test would have revealed to them that the tank car lacked sufficient oxygen to sustain human life. The evidence further revealed that UTC had no company procedures to ensure that its employees were in fact utilizing the testing program. Further, although UTC was aware that Greuner was not actively supervising the employees because of an illness which kept him hospitalized or bedridden, UTC did not replace Greuner, leaving the Whiting Mobile Unit employees virtually unsupervised. R. at 4089.
*891The evidence also supports the Board's finding of deficiencies in UTC's employee training programs. UTC instructed its employees a single time in the use of the Eeo-lyzer, which the Board determined to be insufficient given its complicated operation. R. at 1139, 1141. Although UTC provided its employees with the Ecolyzer's operating instructions, those instructions were located in the multi-volume Tank Car Cleaning Instructions Manual. However, the Whiting Mobile Unit employees were unlikely to access those instructions or to know of their contents because they were only responsible for wiping down the tank cars. Their duties did not include cleaning of the cars. R. at 4252. In fact, UTC employees Bouchee and Anderson testified that they had never used the Tank Car Cleaning Instructions Manual R. at 5720, 6209.
In addition to the testing programs, the evidence also revealed deficiencies in the Ec-olyzer procedures. The Ecolyzer was kept in the Whiting Mobile Unit's main office. As a result, it was necessary for the employees to take the Ecolyzer from the office, test the tank car, return the Eeolyzer to the office, and then return to the tank car to commence wiping it down. This procedure does not permit employees to know the exact level of oxygen or toxicity of the tank cars at the time they enter the tank cars.
Additionally, the evidence supports the Board's finding that UTC did not implement proper communication procedures. Despite UTC's contention that it was the employees' normal practice to work in pairs and to communicate with each other while working inside tank cars, UTC had no policies or procedures requiring its employees to communicate with each other while inside the tank cars or requiring one employee to be posted outside the tank car while the other is inside. UTC also did not provide its employees with devices to enable them to communicate with each other while wiping down the cars. Carter and Anderson's presence inside the tank car demonstrates not only that the employees failed to follow the "normal practice" of communicating with each other, but also that they did not maintain an employee outside of the tank car to summon help if needed.
Further, the evidence supports the Board's determination that UTC failed to implement audit procedures to ensure that its policies and programs were followed. Although UTC had a safety department at its plant in East Chicago and employed the National Loss Control Service Corporation (NATLSCO), an independent safety and health consulting firm, to perform safety audits of its facilities, the procedures used at the Whiting Mobile Unit were neither audited nor supervised at sufficient intervals for UTC to discover the problems at Whiting.4 In fact, a UTC supervisor testified that until the fatal accident, he had never been to the Whiting facility. R. at 5512. Thus, UTC was unaware that its supervisor instructed employees not to use the required procedures or that the employees bad heeded the advice of the supervisor.
Finally, the evidence supports the Board's finding that UTC failed to instruct its employees regarding the hazards of nitrogen and white oil. Although UTC had specification sheets explaining these hazards, the sheets were located in the mobile unit's main office, far from where the employees performed their cleaning of the tank cars. Additionally, the sheets for nitrogen and white oil were located within the Tank Car Cleaning Instructions Manual, which, as we noted above, the employees did not have reason to access and consequently did not use.
In sum, the evidence presented to the Board demonstrates that UTC was well aware of the hazards of entering tank cars and the possibility that the tank ears could be oxygen deficient or contain toxic substances. Nonetheless, the Board found that UTC ignored the safety of its employees by failing to ensure that they safely entered and cleaned the tank cars. Although UTC had programs, policies and procedures for ensur*892ing that employees safely entered tank cars, the record reveals that these efforts were merely illusory in that UTC did not ensure that the programs, policies and procedures were implemented and followed. The evidence, therefore, sufficiently supports the Board's determination that UTC was plainly indifferent to the safety of its employees and its assessment of knowing violations against UTC for 1) failing to identify and evaluate potential hazards, 2) failing to provide and maintain equipment for safe entry, 8) failing to train employees regarding hazards, 4) failing to implement a written permit system and 5) failing to explain the hazards of nitrogen and white oil, We cannot say that the Board's findings were arbitrary or capricious or were unsupported by substantial evidence. We decline UTC's invitation to reweigh the evidence in its favor.
Judgment affirmed.
RILEY, J., concurs. FRIEDLANDER, J., dissents with separate opinion.. IND.CODE §§ 22-8-1.1-1 to 22-8-1.1-51.
. Oral argument was held in Indianapolis on July 9, 1996.
. UTC utilized specification sheets rather than an MSDS to communicate hazards associated with certain chemicals. Specification sheets are patterned after MSDS' and contain all of the information contained on an MSDS as well as some additional information specific to UTC operations.
. UTC states in its brief that NATLSCO audited the mobile units semi-annually, Appellant's Brief at 8, or "at least" annually. Appellant's Brief at 9. However, UTC failed to present evidence regarding the exact frequency of NATLSCO's audits, the methods by which the audits were performed or NATLSCO's findings during its most recent audit of the Whiting Mobile Unit. The absence of these facts bolsters the Board's finding that UTC's safety program was deficient.