There are multiple ways of working with a national lab. CRADA (Cooperative Research and Development Agreement) and WFO (Work for Others) are the two most common routes. Simply put, a CRADA is a research project where the company does some of the work and the lab does some as well. WFO is almost like hiring the lab as a third party contractor. In both cases the company gets to specify what the research is and what the deliverables are to be expected. It is very important to understand that research is by its very nature open-ended, i.e. just because you run an experiment doesn't mean it will be successful, and many such experiments might be necessary to arrive at a successful result, if ever.
A lab is not a product development organization. They are not equipped to do product development for a company. They might be able to build engineering prototypes. They are a research organization, and they solve very difficult technical problems. They have equipment and resources that most small/medium sized companies cannot dream of having.
That is what you are paying for: Unprecedented access to resources in a uniform playing field. General Electric gets the same access as a small company in La Junta, Colorado. If anything, a small company has the advantage because they are small. More about this later.
Everything takes longer when working with the labs. Be prepared for this.
Establishing rapport with the principal investigator, the department manager, and the licensing executive is crucial because you are entirely dependent on them. Visit the lab, see what work has been done, and what they have for results. A face-to-face meeting is worth a hundred emails.
A well-defined statement of work (SOW) is key to making rapid progress and ultimately, success. This is jointly developed between the principal investigator (PI) and the company. The SOW will include timelines, deliverables, reviews, reports, payment terms, etc. The devil is in the details, so understanding the details is important. The lab will then provide an estimate of cost of doing the work. Be prepared to be shell shocked by the overhead rate. National labs are not cheap, but rarely is anything worth having! Having a financial model for the ROI for developing technology and/or product can be helpful in discussions. Opting to do part of the work can reduce costs as well. Startup companies have unique access to sources of funding through the labs themselves.
It is also possible and advisable to sign an 'option to license' at this juncture. It may or may not be an exclusive license, but it preserves your rights as a company and also shows intent and good faith, and is cheap.
Once the CRADA/Work for Others is signed, the real work begins for both parties. That is, after payment is received by the lab. All work is charged to established accounts so until there is money in the project account, work doesn't commence. Likewise, you will be required to maintain a minimum balance. All of which is defined in the agreement. Payment schedules can be negotiated. If there is no money, work stops, and restarting can be slow and time-consuming.
Program management is ultimately the company's responsibility. Clarity of expectations with regard to communication, progress reports is crucial to progress. PIs typically are working on multiple projects, so their attention is divided. They can be temporarily moved to work on projects deemed critical to the lab or the country, such as Fukushima or the Ebola crisis, so your project may get delayed. Be prepared.
Intellectual property developed under a CRADA may be jointly owned if the company was actively involved in its development. The more closely the company is involved in technical issues and in developing solutions the easier it is to be part of intellectual property filings that might arise from the work.
Having clear go/no-go decision points and contingency plans help keep the program on track. Research can be a road with many forks. Being closely involved with the PI ensures the road taken ultimately leads to the desired outcome. Leaving it up to the PI entirely to make those decisions can be problematic. Decisions can have stacked cost implications when it gets to actual product development. The further away you deviate from the main path the easier it is to get lost.
CRADAs can be modified and extended, with or without the addition of money. They can also be terminated and the unused money refunded. The success of the partnership with the lab ultimately lies in staying true to the statement of work, while adapting it to meet the goals of the company, which might evolve as the work progresses.
[This column is the second in a three-part series. Read part one here.]
Mr. Ravishankar is an experienced business leader of technology-based companies and the principal of Operational Solutions, a consulting firm specializing in product innovation and operations. He holds engineering and business degrees from MIT. Reach him at firstname.lastname@example.org.