Otis Elevator v. Intelligent Systems, No. Cv89-0700549s (Jul. 18, 1990) , 1990 Conn. Super. Ct. 649 ( 1990 )

[EDITOR'S NOTE: This case is unpublished as indicated by the issuing court.] MEMORANDUM OF DECISION RE TEMPORARY INJUNCTION Otis Elevator Company (Otis) brings this action seeking, inter alia, a temporary injunction to restrain defendants, United States Elevator, Inc. (United), Central Elevator Company (Central), Intelligent Systems, Inc. (ISI) and David Rado (Rado) from using or disclosing certain trade secrets which Otis claims to own. Those secrets are limited to the seven items listed in plaintiff's post-trial brief of April 3, 1990 on pages 2 and 3.¹

Facts

This case is based on proof of facts. Defendants have provided the court with a huge post-trial appendix and a seventy-two page brief. They have not provided the court any compilation of the facts upon which they believe this case should be decided.

Otis is a large division of United Technologies Corporation (UTC) and it is engaged in the business of manufacturing and installing elevator systems. In 1976 Otis and other UTC divisions, Hamilton Standard and Hamilton Test Systems (collectively "Hamilton"), began the development of a completely computer-controlled elevator system: the Otis Elevonic system (the System). That System placed all the elevator functions under microprocessor computer control, replacing a slower electro-mechanical relays system. This achieved improvements in speed and routing. The System was developed at great cost over a three to four-year period using about 100 employees.

The computer software for the System was developed by a team of software engineers at Hamilton under the supervision of Rado. The System's software contains numerous special "Installation and Maintenance" (IM) routines designed to CT Page 650 permit a service technician to attach the Otis Maintenance Terminal (the OMT)² to the System and thereby communicate with and both control the elevators and perform various tests and diagnostic functions. The OMT permits efficient and fast repair and diagnosis.

In order to access the IM routines, a set of "command codes," consisting of one or more of a series of approximately 40 arbitrary combinations of characters of varying length, must be sent to the Systems controller. When the controller receives those command codes, it performs various operations or tests and sends back a response in an arbitrary format. The Communication Protocols used by the System consist of its particular messages or command codes and the generated responses. These are trade secrets.

Although it is possible to service a Systems elevator without using the built-in IM routines, the ability to access those routines shortens the installation time and permits a service mechanic to be more efficient in diagnosing and correcting problems in the System. Service and maintenance is particularly important in the elevator market, and manufacturers rely heavily on maintenance revenues for their profits.

Recently, Otis has developed a system known as the "Microprocessor Based Group Controller" or "MPGC" system (also referred to as the "ERNI" or "EGC" system). It fits computerized control features to existing, relay based systems. In order to facilitate servicing of MPGC systems, the MPGC software includes some IM routines which use the same Communication Protocols as the System, Thus, the OMT, which was originally designed for the System can also perform the five tests built into the simpler MPGC system.

While developing the System, Otis and Hamilton developed a maintenance terminal capable of accessing the IM routines and interpreting the results. The first such device, which was used during the initial development period, was a bulky computer terminal called the Tektronix 4051 (Tektronix) with special software stored in tape cartridges designed specifically for use in Tektronix terminals. Otis used two Tektronix terminals and three tapes in each of the five Otis geographic regions in the United States and Canada.

In late 1981, Otis began to replace the Tektronix with the OMT. William Aikey (Aikey) had the responsibility for the distribution of the Tektronix tapes and he cautioned Otis' management that those tapes contained software which, in the wrong hands, could be used to derive some of the Communication Protocols. CT Page 651 Aikey was instructed to, and did in fact, recall all of the Tektronix tapes that had been distributed in the field. He held them in his office for several years and then threw them in the trash. However, a few Tektronix tapes containing the IM software were kept at the Otis Engineering Center in Farmington, a secure building with access limited to authorized Otis employees. The tapes were used initially as backups and, in later years, to play a computer game called "Kablooey" which had been included on the tapes by Dave Sorenson, the Otis engineer who produced the production version of the tapes. Joseph Tardif (Tardif), an Otis employee who is now a principal of ISI, took one of those tapes out of the machine in which it was stored, ostensibly to translate Kablooey to run on an IBM PC computer. Without permission, Tardif took the tape home and kept it after leaving Otis' employ. None of the information on that tape was used by ISI in developing its EST.

The Tektronix terminal was used pending the development and production of the OMT. The OMT is a portable computer in an attache case. The software for the OMT was developed over a three-year period by a team of three Hamilton engineers led by David Hardenbrook (Hardenbrook). The Tektronix software cannot be used in an OMT because each uses a different computer language.

In the programs used by the parties to this action, computer software is expressed in two forms: a "source code," which is written in a language intended to be read and understood by human beings; and "object code," which consists of a series of zeros and ones that are understood by the computer.

After being translated into "object code," the software developed by Otis was permanently stored in five EPROM chips in the OMT.

This System is based on a hexadecimal format, i.e. it uses 16 as a base and is a unique code designed by Hamilton. The OMT was designed to provide Otis field personnel convenient access to the IM routines built into the System controllers but it does not disclose to those personnel any information concerning the Communication Protocols or the details of the operation of the IM routines. To do his work in the field the maintenance person simply types test numbers and data on a keyboard whereupon the OMT automatically generates the appropriate command codes, which bear no discernible relationship to the test numbers, and it converts the responses into numbers or a pattern of lights on the screen of the OMT. Using instructions provided in an instruction manual and/or a set of overlays provided with the OMT, neither of which reveals the internal workings of the OMT or the Communication Protocols, the CT Page 652 operator can interpret the results displayed by the lights on the OMT and diagnose any problems in the System.

The OMT was manufactured for Otis by Owl Electronics (Owl), pursuant to an agreement which provided that the device was to be sold exclusively to Otis and that all of the software and other data which Otis provided to Owl was to be preserved as confidential to Otis. Owl built the device from schematics provided by Otis, which bore proprietary legends.

In order to function the OMT needs software. The software is contained in five (EPROMS). The EPROMS contain the object code. Owl used master EPROMS provided by Otis which it kept locked in a safe when not actually in use. Owl never had access to the OMT source code.

From Owl, the OMTs were delivered for internal Otis distribution to the Otis Service Center in Connecticut, where a master file is maintained to keep track of the location and repair record of each OMT by serial number. OMTs are never sold to anyone outside of Otis. But on three occasions Otis extended a lease or license of an OMT to an "outsider," once to a church, once to the United States Government and once to the P.T. Citas Engineering Co., Ltd. in Malaysia. (Citas)

In order to obtain OMTs for use in the field, each Otis region must "purchase" them with the authority of an appropriate management level person. The regions also track the whereabouts of and persons responsible for each OMT. The regions are required to return the OMTs to the service center for repairs or services. No internal parts or software for the OMT are made available to personnel in the field.

In each of the five regions, security for the OMTs is the direct responsibility of the Regional Field Engineer. In general, each OMT is specifically assigned by the Regional Field Engineer to, and becomes the personal responsibility of, a construction superintendent or maintenance supervisor, who is required to sign a receipt for the OMT and accept full responsibility for its security. OMTs left at job sites are required to be kept in locked boxes or cabinets within locked machine rooms when not in use. Although these requirements are not always strictly complied with, Otis installers and mechanics who are given access to OMTs are told that they would be discharged were they to permit an OMT to fall into the hands of anyone outside of Otis.

Otis had in place a program to have all employees execute nondisclosure agreements but it was not totally successful. It also had a program for incorporation of nondisclosure agreement CT Page 653 into all employment applications and purchase orders. The court cannot find whether or not this program was fully implemented. There was also a system to limit access to premises in which development work was carried on and proprietary materials were stored, including video monitoring of doors and parking lots; requiring personnel to carry photo ID badges, and requiring visitors to be logged in and personally escorted while on the premises. This system was in place and working.

Otis had a prohibition against the sale of devices or materials containing proprietary information, such as the OMT, source code listings, Tektronix tapes, and OMT manuals; placing Otis proprietary legends on manuals, prohibiting use or disclosure without Otis' consent; and within Otis, strictly limiting access to source code listings for the Elevonic controllers to senior engineering personnel. All of these prohibitions were in place and appear to have functioned. In addition, while at Otis, employees such as Richard Roberts (Roberts) were required to sign out OMT manuals which were kept in a locked place. If Otis employees released information to outsiders "they would be fired in a heartbeat." United is a direct competitor of Otis in the sale and servicing of elevators.

Central is a wholly-owned subsidiary of United and it services and maintains elevators in the New York metropolitan area.

George C. Tweed (Tweed) was President of United from 1972 until 1979, when he left to join Otis as its Vice President in charge of Technology. In 1982, he was made a Senior Vice President of Otis and placed in charge of all of Otis' North American engineering and manufacturing operations. He held that position until he resigned in October of 1984. However, following his resignation, Otis permitted him to remain on Otis' payroll until January of 1985, which enabled him to collect Otis incentive compensation. Immediately after receiving that additional compensation, he returned to his former employer as President. He is now Vice Chairman and Chief Operating Officer of United's parent company, Cubic Corp.

Shortly after his departure from Otis, Tweed recruited two of Otis' executives to join United. Ferguson, a financial manager who ultimately succeeded Tweed as President of United and Salihi, an engineer who is now in charge of Research and Development (RD) at United.

In 1987, United decided to begin servicing and maintaining microprocessor-based elevators manufactured by others including Otis. At about the same time Tweed began recruiting CT Page 654 another Otis engineer, Rado, who was in charge of all software development at Otis and had served as the principal supervisor for a number of Otis software development projects including, most notably, the development of the Otis Elevonic system.

Rado entered into negotiations with Tweed and Ferguson for a consulting agreement between United and ISI, a corporation Rado had formed with two other Otis engineers in 1986, to conduct a sideline business of installing and maintaining security systems.

Rado executed an agreement on October 12, 1987, under which ISI agreed "to analyze microprocessor-based elevator control systems, as designated from time to time by United, and to supply training, and software as required . . . ." in October 1987 he told Ron Lizzi (Lizzi), an Otis engineer he was recruiting to work for ISI, that he would be developing a device to access and exploit the software features built into the Otis Elevonic system.

In the first half of 1987, the Midwest building business was in a slump and United's business was adversely affected. To increase its business United entered the field of maintaining "foreign" (i.e., competitors') microprocessor-based elevator systems and its Midwest Regional Manager, Dennis De Vos (De Vos) was asked to consider the best way to accomplish that goal. On June 24, 1987, De Vos prepared a report with his recommendations. Among the "foreign" targets was the Otis Elevonic 401 system. De Vos' report noted certain obstacles to servicing Elevonic elevators including:

"The obvious problems to be overcome are board availability and development of a diagnostic/program access device. What I've been able to find out indicates the Otis O.M.T. or equivalent is a virtual necessity to trouble shoot, make routine timing and other adjustments, and so on."

The De Vos report also pointed out the inadequacy of equipment publicly available to perform the functions of an OMT stating:

". . . we have already purchased a device in Dallas for copying and reading chips . . . The device also accesses the program resident on the chip, but this is of little value unless you can interpret what you're seeing."

The report also stated:

"In short, I think it's a move we simply have to CT Page 655 make . . . We need to invest a little seed money in a support structure, whether it be outside vendors or people in house, to basically play with any competitor's equipment we get access to as soon as [we] get it under contract. We need to find a way to do whatever it takes to pull it off a few times and get everyone's confidence up.

Within a few weeks of that memo, Richard Roberts (Roberts) a United employee who had once been employed at Otis, found Larry Parsons (Parsons), another former Otis mechanic who was very knowledgeable about the System, looking for a job at the union hall and De Vos hired Parsons and immediately put him to work at the Saint Paul Medical Center in Dallas (St. Paul).

The St. Paul site had a electro-magnetic, relay-based elevator system which was being maintained by United. However, Otis was then installing a new MPGC system to provide various computerized control functions for that same elevator system. The Otis employee principally involved in the installation project was Steve Jordan (Jordan). He had been assigned an OMT number 1286 to use on that job which was kept locked in a steel "40 box" located in the machine room at the site.

Parsons and Roberts decided that, if they could get Jordan's OMT, they could build a duplicate. Parsons tried to open the 40 box with a duplicate key that he had kept when he left Otis, but his key would not work. Thereafter, Roberts taught Parsons how to pick locks. With that new knowledge, Parsons picked the lock to the 40 box at St. Paul and removed Jordan's OMT.

Roberts and Parsons then went to De Vos and asked for money to support their plan to duplicate the Otis tool. When they started to explain how they planned to accomplish the task, De Vos told them that he did not want to know. However, with the approval of United's Vice President for Branch Operations, Ed Helsely (Helsely), Roberts and Parsons were given authority to spend whatever time was necessary, purchase supplies at company expense and even use a helper named Chris Evans (Evans) to assist them in the project.

Once approval was obtained, Parsons stole OMT 1286 a second time and brought it to Roberts' house. Roberts knew it was an Otis OMT. The device was then taken apart and Roberts took at least 276 photographs of its internal components, including backlit photos of printed circuit boards with the chips removed and detailed close-ups which could be put together to form a large composite. In order to conceal the identity of the device, yellow stickers were pasted to cover CT Page 656 up Otis identifying markings on the back of the faceplate. Parsons also made three copies of the OMT's EPROMS, using a chip copying device which United provided.

During the weeks that followed, Roberts made a set of schematic drawings, using the photographs and, during weekends and Jordan's vacation, the stolen OMT itself. With the assistance of Parsons and Evans (who shopped for parts and did some drawing), Roberts built the "Roberts Tool",³ a device which, when supplied with the Otis software (the EPROMS) taken from OMT 1286, duplicated an OMT's functions, except for a few. Parsons took the Roberts Tool and tested it on the MPGC system at St. Paul in late October. After the test corrections were made on the Roberts Tool and they "touched up the power supply" it was delivered by Roberts to United's Dallas office by the first week in December 1987. Parsons and Evans knew Roberts' purpose in using the purloined tool, making photos and producing the Roberts Tool. Parsons made three copies of the Otis EPROMS at Roberts' home. He left two sets with Roberts and gave one set to De Vos at United's Dallas office.

While Roberts and Parsons were building the Roberts Tool in Dallas, a United engineer, Victor Zhong (Zhong) was given the task of building a similar device at the United's San Diego headquarters. Zhong was provided with two enlarged copies of the photographs of OMT 1286, a set of five programmed EPROMS containing the Otis OMT software and copies of Roberts' schematic drawings. When he needed more information, United sent Zhong to Roberts' house in Dallas on Saturday, August 22, 1987 and Parsons made another trip to the St. Paul site that day to break into the 40 box and bring to Zhong Jordan's OMT for a few hours. Eventually, Zhong was able to make his own working version of an Otis OMT.

While this activity was going forward, Tweed was recruiting Rado, then Otis' senior software engineer, to come to work for United. Rado expressed reluctance to leave Connecticut and United eventually agreed that Rado could render his services through ISI. Under a consulting agreement, formally accepted by Rado on October 12, 1987, United engaged ISI as a "consultant to analyze microprocessor-based elevator control systems, as designated from time to time by [United]." The agreement provided that United reimburse ISI for all of its general administrative expenses, including its rent and utilities, the salaries of all of its personnel (plus a 15% profit factor) and any supplies "and other expenses of [ISI] associated or arising from performance of this Agreement."

During the course of their negotiations, Rado had told Tweed that he did not wish to do any work related to Elevonic CT Page 657 systems for at least a year. However, by the end of October 1987, while Rado was still employed at Otis, he had recruited another Otis software engineer, Ron Lizzi (Lizzi) to work for ISI, telling Lizzi that he would be "writing software to produce a tool for the maintenance of Otis Elevonic elevator controllers." Rado considered the Otis software as confidential.

Tardif, another Otis employee who owned 50% of ISI, produced a set of schematic drawings, dated October 21, 1987, of a device which he labeled an "Elevator Maintenance Terminal." These are copies of drawings done by Roberts. In addition, while still employed by Otis, Mr. Tardif took at least one of the old Tektronix 4051 tapes containing the IM software and kept it at his home.

Rado left Otis' employ on November 30, 1987 and immediately went to Dallas, where he was shown the Roberts Tool. By the end of January 1988, Tardif and Lizzi had also left Otis to join ISI full time.

ISI was involved in several "Otis" projects for United, including: building a duplicate of a communications board for the MPGC system, based on an Otis board which Parsons removed from the St. Paul site; building by Tardif a "second OMT" (the "Tardif Tool") by copying the wiring and the software from the Roberts Tool and using Roberts diagrams, all of which United had sent to ISI; building an "EGC" (another name for the MPGC) emulator, using software copied from the St. Paul system; and writing the software for a new laptop maintenance tool which was to be called the Elevator Service Terminal (EST), in cooperation with Central workers. Lizzi used both the Roberts and Tardif Tools to get the Otis protocols out of the system.

The EPROMS for the Roberts Tool were provided by United. They are Otis' EPROMS. The Tardif Tool's major electronic components and its interface devices are identical to the OMT.

In a status report to Tweed on January 29, 1988, Rado wrote

"We have started digging into the Elevator Service Terminal (EST), formally [sic] called the OMT, based on the information we have received from Dallas. A schedule for the software activity is shown below. As of now, the only other items that will be needed to complete this task is [sic] tapping into an operating EGC system to determine what the communication protocol is. CT Page 658 If we do not get our in house simulator operational by the time software needs this information, we will have to visit a EGC field site with a serial line [protocol] analyzer to capture this information."

ISI recognized that it was working with an Otis tool. From at least January 29, 1988 Rado's status reports refer throughout to "our" OMT and the OMT.

The emulator was not ready in time and in mid-May 1988 Lizzi was sent to Dallas with a protocol analyzer. He met with Roberts and Parsons at St. Paul and, using the Roberts Tool, the protocol analyzer and the MPGC controller at the site, was able to "capture" some of the Communication Protocols.

However, since the MPGC system is not capable of performing all of the tests and diagnostic functions of a true Elevonic system, the information derived was incomplete. As a result on June 9, 1988, Rado wrote: "We will need to capture the specific commands use [sic] by Elevonic over and above those used by EGC. Nancy Gilbert is looking into getting us into the Stamford job next week to capture this additional test information."

A few days later Lizzi took the Tardif Tool to the machine room at the premises at 1055 Washington Boulevard in Stamford, Connecticut (Stamford), which he entered without the knowledge or consent of the building manager, and conducted a series of tests through which he ultimately derived all but a few of the remaining Communication Protocols. Using those Communication Protocols, they completed the software which converts a modified Sharp laptop computer into the United EST.

The development of the Roberts Tool was De Vos' project. De Vos personally approved all of the expenses for the project, including the time of Roberts, Parsons and Evans, and purchases of materials worth at least $10,000.

After a promotion, De Vos assumed responsibility for liaison with ISI and received copies of Rado's status reports showing the efforts to derive the Communication Protocols. He also collected and cataloged United's library of Otis materials, including materials clearly labeled as proprietary. De Vos was the person to whom Rado wrote when Rado wanted to obtain various Otis proprietary materials such as source code listings.

Zhong provided Tardif with additional drawings when CT Page 659 the schematics originally provided to Tardif by U.S. Elevator proved incomplete.

Helsely, a Vice President at United, authorized the funding of the efforts of Roberts and Parsons out of Account 049, a "catch-all" account maintained at United's headquarters. He also initialed Zhong's travel voucher for Zhong's trip to Roberts' home in Dallas.

Tweed, President of United, while the Roberts project was going forward, recruited Rado. He personally approved all travel vouchers in 1987. Tweed received Rado's status reports which described the need to obtain the Communication Protocols.

The court does not believe that defendants have the skills to independently develop a tool that will do what an OMT does. The tools they did make were all direct or indirect copies of plaintiff's OMT.

The defendants have not proven that they either did or could "reverse engineer" the OMT.

The defendants could not "hack" the OMT as a practical matter because it would be almost impossible.

ISI, with the knowledge, consent and active assistance of United personnel, built a device which, directly or indirectly, copied the circuitry and software from an unlawfully obtained OMT. Lizzi, one of the former Otis engineers recruited by Rado for ISI, then hooked the device to a protocol converter and extracted the command codes and data structures used for the various tests. These command codes and data structures were then incorporated into the software for a device which appeared different from the OMT but functioned identically to the OMT. Apparently, access to Otis systems for this initial process was gained under the pretense of maintaining an MPGC.⁴

Through the use of a program called a "disassembler," it is possible to convert object code into a form of source code; but a disassembler is unable to provide the comments, variable names, link maps and other information which is included in the Elevonic source code.

Rado believed that Roberts had built the Roberts Tool by copying an OMT before December 18, 1987. United and ISI worked together to develop a diagnostic tool for Elevonic elevators. CT Page 660

LAW

I. Trade Secrets

"A trade secret may consist of any formula, pattern device or compilation of information which is used in one's business, and which gives him an opportunity to obtain an advantage over competitors who do not know or use it." Allen Mfg. Co. v. Loika, 145 Conn. 509, 515. However, to constitute a trade secret a "substantial element of secrecy must exist, to the extent that there would be difficulty in acquiring the information except by the use of improper means." Town Country House Home Services, Inc. v. Evans, 150 Conn. 314, 318-319.

"Some of the factors to be considered in determining whether given information is a trade secret are: (1) the extent to which the information is known outside the business; (2) the extent to which it is known by employees and others involved in the business; (3) the extent of measures taken by the employer to guard the secrecy of the information; (4) the value of the information to the employer and to his competitors; (5) the amount of effort or money expended by the employer in developing the information; (6) the ease or difficulty with which the information could be properly acquired or duplicated by others. Restatement, 4 Torts 767, comment b." Town Country House Homes Service, Inc. v. Evans, id. 319.

Our Uniform Trade Secrets Act (the Act), C.G.S. 35-50 through 35-58, does not contain a definition of trade secret and thus the court will use the common law definition.

The OMT, its software and its functions are all trade secrets under the common law definition. In addition, they were maintained as trade secrets.

II. Patent

Defendants claim that in plaintiff's patent application for U.S. Patent 7,561,093 (the Patent), it failed to disclose the "best mode contemplated by the inventor of carrying out his invention." From this they argue that the trade secret has been abandoned or at least lost because the "best mode disclosure requirement prevents the simultaneous enjoyment of both patent and trade secrecy protection for a single invention."

This is certainly the law. Picard v. United Aircraft CT Page 661 Corp., 128 F.2d 632, 637 (2d circ), cert. denied 317 U.S. 651;35 U.S.C. § 112. The best mode of using plaintiff's patent is with software. The software plaintiff uses is not patented. Anyone may write and produce any software they wish in connection with the Patent in order for the expected communications to take place. The particular software that plaintiff uses could be replaced by "any person skilled in the art." Plaintiff has made the "best mode" disclosure. The plaintiff's hands are not unclean and it has not acted inequitably.

III. Irreparable Injury and Inadequate Remedy at Law

This court finds that not only has plaintiff been harmed by defendants' actions in stealing and using its trade secrets but that it will be harmed in the future by defendants' continuation of such actions. That harm is irreparable because it cannot be corrected by the payment of money damages. First, money damages for the use of the tools by defendants will be very difficult to compute, and second, the possible worldwide loss of customers would be something from which plaintiff could never recover. To be an adequate remedy at law the remedy must be "one which is specific and adapted to securing the relief sought conveniently, effectively and completely." Pottetti v. Clifford, 146 Conn. 252, 262.

IV. Laches

The defendants have failed to prove any delay by plaintiff that support their claim of laches.

This court cannot find that any delay by plaintiff in bringing this action was inexcusable or that any such delay prejudiced any defendant.

V. Sales and Leases

Defendants claim because plaintiff sold an OMT and leased one or more that it has not kept its "trade secrets" secret and thus there no longer are any trade secrets. Defendants claim that Elevonic controllers are on sale to the public and that the software in the elevator controllers is identical to the software in the OMT.⁵ As a result, they say, that software has been disclosed to the public and is thus no longer a trade secret.

On October 30, 1984 plaintiff leased an OMT to P. T. Citas Engineering Co. Ltd. (Citas). Otis retained title. Under the lease, if the tool were destroyed the remains were to be returned "forthwith to Otis." No person was allowed CT Page 662 to service that OMT other than Otis. Paragraph 8 of that lease forbids disassembling, attempted repair or duplication. The lease is governed by "the internal laws of the Republic of Singapore."

The court finds all of that a reasonable effort at protection of plaintiff's trade secrets in that OMT.

The court cannot find that plaintiff's conduct was not within its pattern of keeping secret its trade secrets.

VI. Burden of Proof

A. The court finds that plaintiff has sustained its burden of proof clearly and convincingly, that defendants had misappropriated its trade secrets as herein described.

B. The defendants have failed to establish any fraud, deception or inequitable conduct on the part of the plaintiff.

The temporary injunction shall issue.

N. O'NEILL, J.