Resume of Ken Hewitt
Manager, WSR Consulting Group, LLC
More than 40 years of diversified experience and expertise in Information Systems. As a user in the manufacturing industry I applied Information Systems to improve product quality and reliability. As a salesman, I educated customers and prospects in how to make best use of and integrate their computer hardware, software with their business and decision flows. During this time, I was able to learn how successful businesses operated.
As a litigation consultant, I was able to use my experiences to drill down and determine what factors were involved in the failure of an Information System and clearly present my findings to the litigation team.
I have worked on 14 closed matters with WSRcg and am currently working on a matter for the Department of Justice with WSRcg. All 14 matters were concluded successfully. I have summarized my involvement in six of these matters below.
- Client 1 was a fortune 500 company that had contracted a software company to develop and install a fault tolerant POS (Point of Sale) system. We designed a test to determine if the system was fault tolerant, It failed and we were able to prove that because of its software architecture, it was not correctable in any reasonable or acceptable timeframe. We also were able to show that it had significant gaps in the area of financial controls.
- Client 2 was a large magazine publisher who had purchased a system to manage their customers, mailings, distribution and promotions. The company who supplied the software had not developed it but had bought it from another publisher who had developed it in house. The system implementation was slow and exposed numerous logical bugs, system inconsistencies, algorithmic errors, and difficulties to audit. The client finally had enough and sued. While at the time of the suit the system was getting the job done, the client had no faith it would be able to continue functioning and the supplier indicated that they would no longer enhance or support the system.
We reviewed the software problem log to see if there were any perceptible trends. One important problem/process we uncovered was that what appeared to be rather simple bugs that should have been easily corrected took many months to fix. We looked in to why it took so long and in the course of our investigation we asked to see the source code. The software was written in BAL, the IBM Basic Assembler Language. When we looked at the BAL listing it became clear to me that it was not source code but a decomplilation of someone’s else’s source code, because, among several other things, it had no comment lines. A programmer would virtually never write a BAL program without putting in a numerous comment lines so that when he had to troubleshoot or change it, he would know were he had to make the changes/fixes That was particularly true of this program as it was a complex algorithm.
We investigated further and the alleged original developer” indicated that this module had been coded by an outside consultant who still had the rights to it. As our client had purchased the source, the vendor had sold something that he did not own. We presented our findings to the Litigation team and the vendor settled.
- Client 3 is a major retail chain who purchased POS Systems from a major POS vendor for installation in all their locations across 15 states. The client did not want to use the bundled systems’ communications design to interface the over 200 stores and a few warehouses that the POS System recommended. The vendor invited our client to the Corporate Headquarters where they were told their “Special Systems Engineers” would explain how they would incorporate their requirement into the system. Two months later they went to the Vendor’s “presentation” they were not presented with a solution but were introduced to a small third party software developer. The vendor assured our client that this software developer would be able to address their concerns. With that guaranty, our client signed the contract.
After a couple of years with no progress by the developer and several POS systems installed at pilot stores, our client called a halt and sued. When we looked into the way the system hardware and software were tightly coupled to the communication software, it appeared to us that the client would never get their concerns addressed unless the vendor undertook a major rewrite of the entire system. Based on our experience in selling and installing systems of this type, it was clear that the vendor’s Special Systems Engineers would have tried to come up with a way to meet the client’s requirements during the two months they had to get ready for the meeting.
That lead us to believe that they had seen the same problems we saw and that during discovery and depositions of the Engineers involved we would get proof. The litigation team wanted to avoid discovery so they asserted the claim that the vendor knew the third party developer would fail but they had hoped that the client would accept the provided communication approach and drop the requirement. The vendor took back all the hardware and refunded to our client all equipment, delivery, software, maintenance and associated costs.
- Client 4, one of the world’s largest law enforcement agencies, contracted with a vendor to deliver an automated fault tolerant field dispatch/reporting system. When the system did not pass the acceptance test, our client cancelled the contract. The vendor sued on the basis that their system would pass if they were allowed to make some minor corrections to the software in a very short period of time. One area where the system failed was that it was not fault tolerant. We interviewed and analyzed the work of the personnel responsible for developing the fault tolerance functionality, their status reports, and testing regimen. We concluded that none of the personnel had ever worked with a fault tolerant system much less developed one and were very unlikely to succeed on their first try.
Reviewing the data flow diagrams that defined their approach to obtaining fault tolerance, we were able to show the litigation team that the opposing teams’ design was unstable due to their error handling approach. This lack of stability would lead to continuous and/or spontaneous crashes, loss of real time emergency communications/use/access to and between the mobile terminals located in each officer’s car and between the cars/officers and the dispatch offices – all demonstrated by their failure during the acceptance test.
- Client 5 is a Federal Agency who was suing a contractor for failure to deliver an automated warehouse system as part of a smart building development effort. The suit had many areas of contention and WSRcg was tasked to deal with several aspects of the material handling system, its robotic automated guided vehicles, its huge computerized conveyor belt systems and much more. Load and stress tests run by the government had to be stopped because the vehicles were “crashing” into each other, and the system could not process the physical loads and drops required -- it was “locked up”.
We reviewed the original simulation and it indicated that the system should not have locked up. So, we next looked at any changes made in the system since the original simulation. We found two major changes. The first change eliminated the requirement for parts needed to be picked out of the bin from the high rise fork lift truck. Originally the pickers were required to have a basket where the operator could remove a part from a storage location and put it in a basket if there was another part that he needed to pick in his area. This would reduce the number of storage bins going on the automated material handling trucks to the central order picking area. The vendor asked this to be dropped and assured our client that the additional bins would not overload the system.
The second requested change was to change the guide paths for the automated material handling trucks from uni-directional to bi-directional in the unsupported and untested hope of dramatically increasing the loads that could be transported in a fixed period of time. Again our client was assured that this would not overload the system. WSRcg had the company that had simulated the original configuration rerun the computerized simulation taking into account the 2 changes. The simulation showed the system going into a “deadly embrace” just as happened in the stress test.
- Client/Litigation 6 involved the investors/developers of an electronic system for the fast food delivery market. The investment group decided not to manufacture or market the system themselves, selling/licensing the manufacturing rights to one company (our client), and the marketing rights to another company. The investor owners of the system kept the selling, delivery and implementation services to themselves. They implemented multiple systems at many locations of the largest franchisee of a major international fast food company. The franchisee claimed the implemented systems’ performance was not acceptable and did not meet the contractual performance requirements. In a complicated and surprising turn of events the investor/implementation company), along with the marketing organization sued our client, the manufacturer of the system. We analyzed numerous logs and operations history documents to get a picture of the magnitude of the problems based on data.
First, we separated the hardware failures from the software failures. Then we broke down the hardware failures by device involved (printer, monitor, cpu, storage drives). Once we had that data, we properly compared the MTBF (Mean Time Between Failure) for each component to the contract required MTBF for each component. We found that the components all met their required individual MTBFs and as there was no MTBF for the whole system, we asserted that the system met specification. When presented with our strong data collection, extracts, analyses and explanations, the marketer dropped his suit against our client and joined in our defense. Other WSRcg staff working on the software/performance problems were able to show that a large percentage of the causes for performance/processing delays were in fact operator errors or a lack of proper training and communication between implementer and user.
I successfully sold mainframe computers, fault tolerant systems and point-of-sales systems for some of the industry’s largest companies for many years. I relied heavily on my knowledge of my products as well as my competitors’ products. Many challenges/issues/lawsuits surrounding systems performance, integration, recovery/restart, and extending a systems useful life all involve a strong background, hands-on experience, and appreciation of how hardware, software, netware and peopleware all work together to support the best solutions to a client’s problems. This knowledge proved to be most helpful on several of the litigation matters I would work on at WSRcg.
One presentation that I created showed the relationship of various hardware architectures to their efficiency in executing different software applications. I was asked to make this presentation to UCLA MBA students on several occasions.
The industries I sold to included:
- scientific research labs
- credit authorization users
- smart building and robotics.
Manufacturing: a few interesting highlights:
I applied statistical techniques developed to improve crop yields to a manufacturing process that had widely varying reject rates and product quality caused by unknown variables, Things like whether the ceiling windows were open or not. Some of the key components were produced in batches but it was very hard to determine if the batch was good or not. If you ran a sample through the process to complete finished products you might have a high rejection rate but that might have been caused by some unknown and uncontrolled variable.
I designed a test where I set aside one batch and called it “normal” production. When a new batch was ready, I selected a random sample from it and from the “normal” batch. Each batch was identified and then mixed and sent through the normal production process to the finished product. The data was collected on both groups then separated and tested for a statistically significant difference. This approach reduced the number of batches rejected to zero for several months representing a significant cost savings. We did reject one batch and when the engineer overseeing the batch process reviewed his records, he found that the batch had been cleaned in fewer small batches. He changed the cleaning procedure to use more small batches and we were able to show that the change improved the product quality.
I developed and managed a failure analysis department for a large government contractor. The results were so beneficial that the government added a requirement for a failure analysis function on add-on contracts.
|WSR Consulting Group, LLC, Los Angeles, California
||1996 - Present
Senior Litigation Consultant
Worked on 14 completed matters and one open matter.
| Computer Sales.
|| 1972 - 1995
Met sales objectives in all years
Worked as a Quality Control (QC) Engineer, QC Manager, QC Director
| MBA (Masters in Business Administration)
Cal State University Northridge, Northridge, CA
| BS Math (Baccalaureate degree)
University of St Thomas, St Paul, MN
Illinois Benedictine, Lisle, IL
Selected to compete in the prestigious Lowell Putnam Mathematical Competition 1958
Named to Who’s Who in Mathematics, 1959