3 mins

Automation systems do not always incorporate AI. However, some automation systems use AI to process data. For example, a common application of AI in automation is to detect objects in images. Through AI, a machine can be trained to recognise and label objects in images (e.g. alphabets and words on a scanned document, otherwise known as ‘optical character recognition’ (OCR), or pedestrians from a car’s dashboard camera). These inputs can then be fed into an automation system to generate an appropriate response (e.g. extracting payment dates and amounts from paper invoices, or steering automatically to avoid a pedestrian).

Automation facilitates the reduction of human effort required for repetitive tasks or processes, creating capacity for individual experts to engage in more complex and high-value work. Automation also enables these processes to be run on-demand, facilitating faster delivery. Combined, these advantages allow automation to play a vital role in effectively and efficiently gathering data from multiple sources on a regular basis. The outputs generated by automation would also be in a structured format, so as to facilitate further automation or data analysis to gain insights that will inform decision-making processes.

Low-code and no-code platforms can be a cost-effective way of deploying automation, without the need to involve specialist computer programming expertise. To help reduce initial costs, automation can also be tested on discrete and minor processes to gauge value and efficiency. However, the successful design of any automated workflow often depends on all stakeholders contributing their expertise to the process. As such, sufficient upfront capacity from the relevant teams should be allocated into the automation design process.

Beyond a cost-benefit analysis e.g. will automation bring sufficient savings when weighed against initial investment, consider whether the underlying process is suitable for automation. If human judgment or empathy is a key aspect of the relevant task e.g. responding to complaints or engaging with clients in distress, then automation is unlikely to be suitable. Data quality is also a consideration. If the necessary data is not sufficiently structured it may require considerable manual preparation before automation is possible. If this is the case, consider starting automation further upstream in the workflow to prepare the input data necessary for automating subsequent processes.

It is important that a team retains its know-how relevant to the task being automated, even after deployment. To start, successful automation relies on high quality input data and team members will need to ensure the input data is sufficiently prepared with a view to what is required by the automated processes. Moreover, the retention of this know-how serves as a contingency, whilst allowing the team to successfully monitor the system and deviate from it when needed. Moving forward, up-to-date knowledge allows a team to constantly evaluate and improve a system when needed, reflecting any necessary changes in the automated workflow.

Thank you for signing up for LawtechUK news and updates

You will now be notified of all our latest updates.

Thank you for contacting LawtechUK

We will get back to you as soon as we can.

Thank you for your interest in the Open Legal Data project.

The LawtechUK team will be in touch with you shortly.

Thank you for signing up for LawtechUK news and updates

You will now be notified of all our latest updates.