The primary aim of our technology assessment work is to identify best practices as well as potential problems with LED system and component quality and performance and to sensitize manufacturers, vendors, and other players to the need to take this into account. We expect that our work will inspire others to set up full-scale quality testing and labeling programs that will evaluate specific products and place those results in the public domain. We also seek to inspire large purchasers to establish performance specifications and verify that products meet those specifications before they order or make payment. Part of our contribution will be in developing the underlying test protocols, and absolute benchmarks that products could attain. We believe this is a worthwhile pursuit, and it has not been previously done in the case of white LED lighting systems for off-grid applications. When we are able, we will send results for individual products back to individual manufacturers or vendors on a confidential basis. While we realize that many parties may request that we associate brand names with our test results, following are the practical reasons that this is problematic:

  • Comprehensive test procedures have yet to be developed and adopted. Until they exist, it would create confusion if different groups promulgate non-standardized product evaluations.
  • We are a public research lab, not an accredited commercial testing lab. There are standard conditions that should be applied for many of the tests—e.g. 25C around the LEDs—for which we are not always equipped. In some cases, the equipment required to do the tests properly would be prohibitively expensive to acquire.
  • We have noted significant variations not only among individual "identical" LEDs but also for "identical" systems, and are not necessarily able to do large enough multiple-unit testing to obtain statistically significant results or to identify all important "outlier" results and problems such as failure rates.
  • By the time we release any particular result, the products may have evolved. Manufacturers could understandably request that we re-test a growing number of systems and components with great regularity, which is something for which we may not have resources.
  • Manufacturers of products that receive "good grades" (using their judgement of what constitutes "good", not necessarily ours) may inappropriately assert that we endorse or otherwise certify their products. It will be up to others to establish labeling or rating systems for these products.