AI and other computer models are worse than useless with bad data
John Reid
Austraffic, Victoria
This presentation was delivered at the 2021 Online Conference Series and until October 2022 is only available to registered delegates and Content Access Pass holders via Interchange. For information on accessing this and other presentations please review the Content Access Pass options.
ABSTRACT
The impact of COVID has been huge, but it is also being used as an excuse to cover over some long term, systemic problems. The pandemic has highlighted how ill-prepared we are for assessing and tackling different situations. For too long we have glossed over data being badly collected and then misused.
COVID has shown that without the right, comprehensive data we can destroy the lives of many people while we wallow in unsubstantiated opinions. New technologies can now analysis numbers at a faster rate and in more complex (read hidden) ways, thereby increasing the rush to inappropriate solutions.
A recent article in the Financial Review said in part:
"With so much data flowing freely across the internet, and with so many technologies evolving that can easily manipulate it, the time has come to think about the quality of the data, and how it is secured and managed, experts say".
Practical examples will be used to show how:
- Some history of a disregard for good data and a belief in clever technology
- How we have measured existing movements but not underlying reasons why people travel
- We have been misled by the use of averages
- How data that is just one snapshot in time is used to defining the whole, on-going situation
- How we have measured one aspect without looking at other consequences
- How clever companies are not afraid to fund the collection of good data.
The Way Forward
Generalities about how we must “try harder” are not enough. We have to stand our ground on demanding the right resources to create the right foundation. Governments must also be prepared to set mandatory standards, not just blame market forces.
We must start the change now. Governments, private industry and academia all have a role to play in how we manage the following:
- What is being collected;
- How is it being collected
- Make relevance and accuracy, not price, the dominant factors
- Audits of the veracity of the data;
- Resources on how we store the data; and
- Managing how it is being used?
Author
John Reid | Austraffic
John has over 34 years experience running Austraffic, a transport data collection business, and has made major contributions to the development of the profession through Institutes and academia.
Having begun his career in local government John then started his own transport survey company Austraffic, 34 years ago. With branches in all states of Australia and in NZ and his organisation has a reputation for pursuing quality not just churning out numbers.
John has a passion for progressing the transport planning and management industry and the benefits it can give to the community.
John has given time and financial support to industry groups and associations especially in encouraging young people and the development and retention of profession skills and knowledge throughout all the organisations that are part of providing transport systems.
Recently, John has published and presented ideas and suggestions on how we look at and understand big data. This is not just a case of being fascinated by new technology but also incorporates the history and wisdom that has built up from the very beginning of our profession.