A portable emissions measurement system (PEMS) is essentially a lightweight ‘laboratory’ that is used to test and/or assess mobile source emissions (i.e. cars, trucks, buses, construction equipment, generators, trains, cranes, etc.) for the purposes of compliance, regulation, or decision-making. Early examples of vehicle emissions equipment were developed and marketed by Warren Spring Laboratory UK during the early 1990s. This equipment was used to measure on-road emissions as part of the UK Environment Research Programme. Governmental entities like United States Environmental Protection Agency (USEPA), European Union, as well as various states and private sector entities have begun to utilize PEMS in order to reduce both the costs and time involved in making mobile emissions decisions. Various state, federal, and international agencies began referring to this shorthand term in the early 2000s, and the nickname became part of industry parlance.
Since the mid-19th century, dynamometers (or "dyno" for short) have been used to measure torque and rotational speed (measured in rpm) from which power produced by an engine, motor or other rotating can then be calculated. A chassis dynamometer measures power from the engine through the wheels. The vehicle is parked on rollers which the car then turns and the output is measured. These dynos can be fixed or portable. Because of frictional and mechanical losses in the various drivetrain components, the measured horsepower is generally 15–20 percent less than the brake horsepower measured at the crankshaft or flywheel on an engine dynamometer. Historically though, dynamometer emission tests are very expensive, and have usually involved removing fleet vehicles from service for a long period of time. Also, the data derived from such testing is not representative of “real world” driving conditions, not least due to numerous and ample flexibilities in laboratory test procedures, particularly in the European Union's "New European Driving Dycle (NEDC)". Beyond that, this test method cannot be deemed as representative, especially due to the relatively low amount of repeatable tests at such a facility.