Inspired by autonomous vehicles classification scale, we propose here a scale for expert augmentation and automation software.
In non-autonomous vehicle; the driver has sole control of the vehicle. On his blog, Abdul Dremaly summarizes well the five-level classification scale for autonomous vehicles proposed by The Society of Automotive Engineers (SAE). In order to help automotive engineers, governments, and insurers get a better handle on this new technology, SAE defined the five levels of automobile autonomy:
A single function is automated, but does not necessarily use information about the driving environment. A car operating under simple cruise control would qualify as Level 1.
Acceleration, deceleration and steering are automated, and use sensory input from the environment to make decisions. Modern cars that have cruise control and automated lane maintenance or collision-avoidance braking would fall into this category. The driver is still ultimately responsible for the safe operation of the car.
In this level, all safety functions are automated, but the driver is still needed to take over in an emergency that the car can’t handle. Tesla vehicles with the “autopilot” feature engaged are examples. This is the most controversial level because it requires the human driver to remain alert and focused on the task of driving, even though the car is doing most of the work.
Fully autonomous level, where the car handles all driving decisions with no input from a human at all. Level 4 cars are limited to a specific set of driving scenarios, such as city, suburban and highway driving.
Fully autonomous level, can handle any driving scenario, including off-road operation.
In Expert Augmentation Software:
Here’s a scale for Expert Augmentation and Automation Software (EAAS). EAAS is a type of software platform where subject matter experts perform cognitive tasks such as summarizing information, linking concepts or indexing documents with the help of automation tools such as machine learning (ML) or robotic software automation (RPA). These types of software programs and systems are said to be taking over the work of semi-skilled white-collar workers in the knowledge economy.
A simple function is automated and does not need to be re-trained or changed over time. This applies to closed-world tasks (e.g., writing up a meteo report based on a small set of numeric values), activities where rules are fixed and can’t change (e.g., sorting images between black and white, grayscale or color).
A more complex function is automated using rules or ML and must be trained or updated from time to time to maintain its accuracy. As time goes and world changes, new examples must be taken into account by the system to perform well. Training must be performed by EAAS engineer, either by encoding knowledge into the system or by applying machine learning on a continuously growing dataset.
In this level, a more complex function is automated with ML plus an automated feedback loop. As time goes and world changes, the system is automatically retrained and updated without requiring an EAAS engineer to oversee the operations. Level 3 systems are at the mercy of concept drifts, change in the task accomplished or inconsistant work from experts. Upon detection, such problems have to be solved by EAAS engineers.
Fully autonomous level, where EAAS system can solve a problem in the first place but also diagnose inconsistencies and adjust target for concept drifts automatically.
Fully autonomous level, where EAAS system can solve a problem in the first place, diagnose and adjust target concept drifts automatically and improve system performance by automatically covering larger problems or discovering and implementing higher-accuracy techniques.