Robots are very useful for performing difficult, complex, dangerous, and/or repetitive tasks, which is why they have infiltrated industry. In science labs though, robots are not always asked to perform repetitive tasks, which is a problem for the scientists. The programming languages to instruct a robot are not developed with a biologist in mind and their flexible workload. Researchers at Join BioEnergy Institute, which includes Berkeley Lab, have developed a new programming language for those with a greater understanding of biology than software to use.
Called PaR-PaR, for Programming a Robot, the language recognizes physical laboratory objects as virtual objects with specific properties. Specific actions can then be performed on these objects in sequences that can then be converted into PaR-PaR commands. These commands are then processed through a database to optimize them for the specific robot platform being used. As no platform directly uses PaR-PaR, this conversion is needed, but the conversion process can be used to translate the PaR-PaR commands to any platform desired.
Typically if a robot needed to be programmed and the scientists could not do it themselves, they would hire a professional programmer to do it for them or use a vendor supplied interface with limited capabilities. With easier access to a high-level programming language, more experiments can be done at less cost and with greater speed.