University-Small Business Collaboration:

Case Studies in Agile Product Development

James S. Burns, Ph.D.

Department of Mechanical Engineering
San Diego State University
San Diego, CA 92182-1323
(619) 594-6076
(619) 594-6005 FAX
jburns@mail.sdsu.edu

 

ABSTRACT

San Diego State University has embraced Agile Product Development as a means of generating value for its customers. Among these are a predominant number of startup and under-funded customers with potentially worthy ideas, but little access to the competitive advantages available to larger firms in the form of enterprise management software, rapid prototyping and expert consulting. This paper discusses, in detail, three typical, though diverse, product development processes (PRPs) incubated at SDSU during the past year as well as a number of other projects. Emphasis will be given in the paper to 1) SDSU's agile manufacturing vision and capabilities, 2) customer involvement, 3) student involvement as both customer and worker, 4) faculty involvement, 5) pitfalls and lessons learned.

 

BACKGROUND

San Diego State University has consistently led the California State University system in annual research funding. The larger share of that does not pass through the College of Engineering, however. The College of Engineering was separated from the College of Sciences over a decade ago, but continues to be the smallest of the SDSU colleges. Beginning in the late 1980's, a funding crisis in California state government led in part to almost five years of decline in the college. The Mechanical Engineering Department full-time faculty decreased by 40%, equipment and facilities budgeting suffered, and faculty class-loads increased without commensurate increases in salary.
The community at large also suffered during this period; San Diego's status as an aerospace Mecca and high-tech job market underwent startling transformation. Biotechnology and telecommunications upstarts bloomed where defense giants once flourished. Mechanical engineering skills were redirected, through down-sizing and plant-closure, away from large companies and into smaller companies that were often outside the aerospace sector. 1993 US Department of Labor employment statistics noted double-digit employment drops in the durable goods categories associated with aerospace/defense. The largest manufacturing segment increase in San Diego in the last few years was for plastic and rubber goods manufacturing. This employment segment services biotechnology, telecommunications and several other major growth industries, but usually indirectly e.g. from the supplier side. Because of this indirect association to growth industries, the author found little initial enthusiasm for the use of plastics and composites as the focal technology segment for a new manufacturing option within the Mechanical Engineering Department and its nexus of related activity named the Facility for Applied Manufacturing Enterprise (FAME).
The key factors in the final decision to focus FAME on net-shape processing of advanced materials were the author's extensive professional and academic experience in application of composite materials, a college-level desire to build academic reputation in hybrid-electric vehicle technologies which rely on lightweight structures, and the author's early solicitation from private industry of over $500,000 in relevant manufacturing equipment donations. FAME's larger, non-technical mission, though, is to promote academic reform that supports agile manufacturing, of which net-shape, advanced material technologies are a big part.

 

CUSTOMER INVOLVEMENT

Identification of SDSU's customers should be as easy as reading the College of Engineering's mission statement. An older mission statement, one that has been recently redrafted, is presented on the college's web page as:

"The objective of the engineering program at San Diego State is to provide the intellectual and physical environment best calculated to encourage students to develop their capacities toward a successful career in the profession of engineering, knowing the need for engineers to maintain a professional proficiency in a rapidly changing technology and advancing state of the art. Moreover, the effective development and application of technology depends on responsible judgments by professionals cognizant of the total needs of society and how technology affects people. Thus, the engineering graduate should have the academic background necessary for personal and professional growth. These goals determine the content of the undergraduate engineering program".

Statements of this type typify academic institutions and accentuate the chasm existing between the expectations of the consumers of the value-producing process of engineering education, and the purveyors of said value. The author takes exception to the older goal statement because 1) the definition of customer is vague, 2) features and benefits of the implied value-production process are vague, 3) the over-arching objective is to produce an "environment", not to deliver quality students to the employment market (e.g. the process is not product-driven). In each of the manufacturing electives taught at SDSU, the author strongly encourages customer-defined, value producing student projects and learning-by-doing through hands-on assignments and class projects with tangible deliverables.
SDSU's new Dean of the College of Engineering and his new draft mission statement focus on (among other things):

"...high quality, practice-based undergraduate engineering education...and a Total Quality Design paradigm throughout the curriculum".

"...high quality, graduate-level education to contribute to the economic development of the region".

"...to act as a brain trust for local industry".
"...to provide life-long learning opportunities...".

This new focus, along with draft features of a solution designed to accomplish these goals through measurement and continuous improvement, go a long way toward reasserting what the author believes is a responsive roll for academia.

The Mechanical Engineering Department has several formal avenues for customer involvement. The first is through its well established senior design program. To quote from the SDSU ME web page: http://kahuna.sdsu.edu/engineering/mechanical/

"San Diego State University was one of the first universities to recognize the value of Senior Projects as an excellent means to teach Mechanical Engineering students a professional approach to functioning as an engineer. Our Senior Project course started in 1957 and has grown significantly in numbers and complexity and industrial support. We believe, and employers of our graduates support us, that participation in the Senior Project course prepares the graduate to deal effectively with loosely structured design problems. In such courses, students tackle open-ended engineering problems whose solutions require a synthesis of design know-how, judgment, technical skills, analysis, creativity and innovation. They also find out that schedules, dollars, procurement time, machining skills, purchasing know-how are sometimes as important as theory. Today, Senior Projects is a two semester sequence of courses taken in the senior year. Students are required to design, build and test an engineering system or perform a professional task of equal rigor. Many of the student projects are sponsored by industrial concerns through a departmental "industry-sponsorship" program called Industry Sponsored Senior Projects (ISSP). Our ISSP Program is designed to provide: - A realistic experience in the practice of engineering problem solving for our undergraduates, and - An opportunity for industry to obtain engineering studies directly related to their needs through sponsored projects with appropriate deadlines and with considerable savings".

The above description of Senior Projects allows for both externally funded efforts and internally funded efforts with external customers. There is generally a bias in the minds of students toward external-customer projects with external funding due to the associated brag factor and the possibility of follow-on employment implied by external sponsorship. That bias reverses if the student feels that the external project is merely an exploitation of their "cheap" labor or doesn't contain elements that lead to personal and professional growth. Faculty have a bias toward diverting the best students toward internal projects. Cooperative projects between faculty and external sponsors seem to offer the student the greatest share of customer involvement provided the student is equally enthusiastic about the project and considers him/herself as a customer.

An example of a project in which the student considers him/herself a secondary customer is the project entitled "Fluoride Tray with Saliva Ejector" that began in June of 1996. The sponsor was a dental hygienist who had heard about SDSU's rapid prototyping capabilities "through the grape vine". It should be noted that direct promotion of FAME via trade shows and mailings has, thus far, proven too costly and so "grape vine" methods have been used exclusively to date. The sponsor was seeking help in the development of an in-mouth fluoride-holding tray that would also contain integral saliva evacuation channels. The student had sought out the author, prior to the inception of the project, and displayed an aptitude and personal interest for paid summer work using parametric CAD. The author's interest was to study this one-on-one product development process. The sponsor, the student and the faculty member formed the design team.

Existing fluoride application trays require a separate saliva ejector, due to increased salivation caused by the fluoride paste. With a conventional tray, a large quantity of saliva pools beneath the tongue, and saliva from the parotid glands will flow over the bottom lip. The traditional method for removing this saliva with a separate ejector is to move it around the areas of the mouth where saliva gathers. The sponsor's idea would remedy this situation by providing the tray with saliva-evacuation channels that lead to one exit port into which a standard vacuum-assisted saliva ejector would be inserted.

A functional prototype was delivered to SDSU by the sponsor. It consisted of two separate vacuum-formed molds of low-density polyethylene sheeting bonded together with hot glue. To attach the two molds required painstaking alignment and gluing. Neither of these operations were desired by our sponsor. The design team's first concept, shown in Figure 1, integrated the two separate pieces of material as well as the gluing operation. The concept consisted of a single vacuum molding with a feature that flips over to create an evacuation tunnel for the saliva. The main section of the part was vacuum-formed over a male mold segment. The feature that flipped over during assembly was formed against a female mold insert. The "flap" was secured by placing 5/8" shrink-fit tubing over the two halves of the salivary exit where the saliva ejector was inserted.


Figure 1. First design concept.

To create the mold for this project, Pro/Engineer solid modeling software was used. The student spent approximately eighty hours learning the software before feeling confident enough to begin designing the part. Due to the complex geometry of the device, approximately forty hours were necessary to create a solid model, which consisted of both the male and female molds. SLA files was then created for both mold segments and transferred to the stereolithography machine where the segments were fabricated.

The SL mold segments were attached to a piece of hardwood, as shown in Figure 2. The female mold segment was positioned on a plane beneath the male mold by milling out part of the wood. The molds and their mount were placed in a vacuum-forming machine. Low-density polyethylene sheeting (0.200 in.) was chosen as the material to form with due to its flexibility and recyclable characteristics. Once the sheet was formed approximately eighty minutes (!) of trimming was required to produce a finished part. The team suggested that the sponsor purchase a knife-edge cutting die for this purpose. Finally, the shrink tubing was applied and a heat source was introduced to shrink the tubing to a desired fit.

Figure 2. SL mold segments shown attached to hardwood base.

Upon inspection of the finished part, as shown in Figure 3, the team noted difficulty in repeatably locating the "flap" prior to assembly. Due to the choice of two-hundredths inch sheeting, the flap became very thin and flimsy as it stretched to form inside the female mold. After several unsuccessful attempts to alleviate this problem, the team decided on a design change.

Figure 3. Device before and after design change.

Instead of having the "flap" flip underneath the male mold to cover the channel, the new idea would have it flip over the male mold. The channel would be placed on top of and the flap pulled over the male mold thus forming a seal despite the thin material. Saliva could enter the channel through strategic cuts (Figure 3) in the flap on each side near the area where saliva generally pools under the tongue. The two halves of the "flap" and male mold which were connected with shrink-fit tubing had to be manipulated slightly for the new design, but worked effectively.

The final product has been presented to our sponsor for prearranged off-site clinical trials. In both design variants, the removal of the parotid gland saliva was by means of holes drilled around the neck of the evacuation tube. These holes represented undesirable secondary operations, however, and further brainstorming will be done to try to create a channel in the male mold to allow the saliva to run into the evacuation tunnel. Furthermore, due to trimming of the polyethylene, sharp edges are created which produce discomfort in the gums. The sponsor has suggested a lining material for further studies.

 

STUDENT INVOLVEMENT AS BOTH CUSTOMER AND WORKER

Another somewhat different PRP example is titled "Mine Reconnaissance Underwater Vehicle (MRUV) Development". In this example, the sponsor (a retired Navy officer with his own small-business) received a Phase II SBIR award to manufacture a mine countermeasure vehicle. The sponsor-now-student enrolled in manufacturing courses and negotiated for class projects related to his PRP-related interests. The first of these projects was for the detail packaging design (but not analysis) of a housing, shown in Figure 4, for a number of vehicle subsystems located in the fore-body of the vehicle.

The MRUV is a modular mine reconnaissance and neutralization system in which sonar, range-gated lidar, GPS and acoustic navigation components are arranged in a roughly 5" diameter by 5 ft. long package weighing less than 40 lb. It is man-portable and designed for deployment in a variety of modes. The MRUV can be operated by divers or, through a fiber-optic tether, by remote personnel. Because the student/customer was formerly part of several end-customer groups, he brings near-complete understanding of all customer wants to the PRP team.


Figure 4. MRUV Fore-body Housing holds imaging and display systems.

The old adage about price, schedule or quality (pick any two!) had forced the customer/student to embrace solid modeling-rapid prototyping to produce the fore-body housing rather than to choose more traditional casting (too long) or machining (too expensive) options. Stereolithography provided prototype functional assessment through sea-going trials of the device. Quality remains a concern because the device was not analyzed in any robust engineering sense by the student. The student learned the basics of ProENGINEER during his academic projects and spent many additional hours polishing those skills. The student also took more than fair advantage of expert design and machining advice, machine use and other resources available within FAME. Several part-time employment opportunities were provided by the customer/student to other students that possessed design and fabrication skills.

A second component for the MRUV is the canister partition for a fiber-optic payout mechanism. This component is shown in Figure 5. This component and the previous one were both lighter in near-net-shape SLA form than in the metal concept designs. Weight is a critical concern in this PRP because savings in structural components permit greater on-board battery capacity and resulting greater range.



Figure 5. Fiber-optic Payout Device

Another PRP in which the sponsor or primary customer is also the student worker, is one titled "Classic Automobile Lighting Accessories". In this PRP, the student proposed fabrication of a prototype hard-to-find lens piece for a vintage automobile. The student's familiarity with the enthusiasts market for such items and with the process of clear plastic casting techniques was utilized by the author to create a win-win scenario for SDSU and the student. The student designed and fabricated a vacuum casting apparatus, similar in nature to those used by establish rapid prototyping firms to create multiple copies of a given RP part. The student specified and purchased materials for a proof-of-concept, silicone-molded urethane part production system as part of an internal Senior Design project. The student was subsequently allowed access to the silicone casting molds for the production of a batch of parts for his own profit.

 

FACULTY INVOLVEMENT

The role of the faculty member in agile product development within a university setting depends on the level of support available through other university resources. The faculty member, due to his/her sometimes over-estimated credibility in technical matters, is often required for all marketing and sales phases of a PRP. This is especially true if the amount of money involved in the proposed PRP is small, because the university's fiscal entity demands a high price for their participation in any contracted effort. In addition, student workers require considerable time investment from faculty due to under-developed judgment skills, incomplete domain-specific knowledge and the prevalent preference held by students for inaction over mistake making. Universities with substantial support staff and overhead budgets are better prepared to deal with the short-term technician support and miscellaneous budget needs associated with agile manufacturing efforts that occur for profit and on a schedule.

Since PRP efforts in universities require students if they are to fulfill educational charters, student turnover and the associated loss of expertise is a nettlesome issue. Training of technicians for key skill retention is not an answer in most situations because either the technician is largely incapable of grasping the non-traditional skills of agile manufacturing or any technician trainable in these skill areas can be expected to leave for higher salaries elsewhere. Graduate students are a partial answer to this problem; they provide multi-year continuity essential for skill retention and timely transmission. Unfortunately, graduate work in PRP is usually confined to more theoretical aspects such as Information and Management Systems. Research that EVENTUALLY enables external PRP is common. Hands-on PRP that proceeds for the benefit of market driven customers is scarce in academia.

Faculty sponsored projects are common at SDSU. They generally don't provide the student with the same level of customer contact, enthusiasm or complete PRP exposure simply because the faculty projects are less market-driven than those of external customers. An example of an author-sponsored PRP is titled "LabVOX: A LabVIEW mechatronic demonstration box and project kit" and is shown in Figure 6. The author's initial efforts to teach LabVIEW programming skills to undergraduates was hampered by the absence of a compact hardware device through which the student could interact with the software and hardware of LabVIEW. The student PRP team negotiated with the author/customer about feature and benefit compromises, struggled with the constraints offered by LabVIEW, budget, and schedule. The prototype component layout was presented first as a series of Pro/ENGINEER renderings and then constructed using stereolithography. Off the shelf component parts such as LEDs switches and pots were installed, and the internal circuit-board design was finalized. The enclosure was assembled sans internal electronic hardware and the project was passed to a "subcontracted" student for finishing.

Figure 6. LabVOX prototype.

 

PITFALLS AND LESSONS LEARNED

Ease of assembly through reduced part count and design-for-assembly was deemed by the dental device PRP team to be important to the economical production of a competitive dental device. The design tool chosen gave the PRP team the agility to accommodate changes quickly by manipulating three-dimensional modeling designs and transferring the new models to SLA files. New molds, in a range of sizes to fit the mouths of patients of different ages could be created in several hours using stereolithography. In summary, the dental device PRP, in the words of the student,

"... provided a valuable learning experience for the involved student. An extensive amount of seat time was necessary to begin with the three-dimensional solid modeling. Over two-hundred fifty hours were spent learning the software and creating solid models. Advanced solid creating commands, which are not covered in introductory books on Pro/ENGINEER software, were implemented for the complex geometry. Upon successful creation of this project, further work on solid modeling design will be accelerated for the student due to the advanced experience. Also acquired was the ability to visualize how a mold will be necessarily designed to provide an acceptable vacuum-formed part. Introduction to stereolithography for the use of mold fabrication was valuable for future design projects involving plastics-forming processes. However, most importantly, the experience of communicating effectively with a customer provided the most benefit for the student. Without feedback from the customer and scheduled design meetings, the student would have had more difficulty developing the product to a point where it would carry out it's intended function. Moreover, due to feedback from the customer's expertise on saliva evacuation and fluoride application trays, product development was accelerated and the design intent was more clearly envisioned by the student".

The team lacked the budget and domain knowledge required to solve all the production problems to its satisfaction. The device did received US Patent Number 5,513,986. Further effort will probably be expended toward an injection molding solution to the fit, trimming and discomfort problems that were unaddressable with vacuum forming production techniques.

The MRUV PRP clearly educated the author on the need for a well-communicated policy on the monopolization of PRP resources by customers who are also students. The access such a student has to resources, human and otherwise, may lead to abuses of privilege and faith. This project is a success for the sponsor, but a pyrrhic victory for the author. Success for the author's overall efforts to build a PRP-based manufacturing program in an academic setting seems to hinge on low-volume, low-investment, relatively high-return efforts and is compromised by high-volume, (unintentionally) heavily subsidized PRPs for sponsors who could compensate the author with intangibles such as those provided by networking or up-front payment, but have been slow in doing so. The attitude that a sponsor's money and/ or prestige are sufficient contributions to the overall satisfaction of the team (a worthy secondary goal of any PRP) is a common team- and organization-disabling fault in the industry-university partnership displayed in this case. The team must be able to continue after the sponsor has departed.

Ayn Rand's concept of enlightened self-interest can be used effectively in the university environment to stimulate student participation. The Automobile Lens PRP was a success except in the area of documentation. The ability to pass on the techniques learned by the first team to a later team were compromised by the personal weaknesses of the team members in written communication. Hands-on types tend to under emphasize writing and to wander off when the hardware is fabricated. Grades are an ineffective hold on such teams, but integrity is commonly enough to force complete performance. A video run-through of the production procedure is pending.