Hacker Newsnew | past | comments | ask | show | jobs | submit | farmerdee's commentslogin

This tool does two things;

1. A neural network predicts probable faults in used vehicles based on the vehicle's make, model, mileage and age. The predictions are derived from previous MOT failures for vehicles of a similar condition (MOTs are an annual vehicle test in the UK to assure roadworthiness). You can enter in those details at https://avoidalemon.com

2. (Of limited use outside the UK) The tool provides a visual representation of mileage over time for anybody who has a UK registered vehicle. The odometer reading over time is plotted on a graph which makes it is easier to see inflections which are indicative of changes in usage/ownership. If the vehicle has had the mileage clocked then this would also be more apparent. You can enter any registration at https://avoidalemon.com/car-report .

This is not a replacement for mechanical expertise or even for consulting suitable motoring forums. What I am hoping this tool does is provide a steer on where issues may be present based on previous MOT failures.

For those of you interested in the underlying data, the UK Government has open sourced a lot of data pertaining to vehicles. You can read more about it here - https://dvsa.github.io/mot-history-api-documentation/ , and here - https://developer-portal.driver-vehicle-licensing.api.gov.uk... .

Fair warning should you want to try this yourself – the data, particularly the MOT API, is very messy and required a significant preparation. Mechanics enter data into the national database mostly by freetext and there is little in the way of standardisation. For example, the manufacturer “Peugeot” was spelled in at least 10 different ways and, rather hilariously, the MOT datasheet is occasionally used to insult the vehicle owner - https://www.mirror.co.uk/news/uk-news/two-incredibly-rude-mo...

Let me know what you think!


Thanks for giving it a whirl! Having some kind of subscription and monitoring service is definitely an option in terms of where to take it. I'm torn on whether my time is best spent doing that or adding features and improving the UI. Good to know its useful!


Hi all,

Please see my new tool, PrivacyTrail. Feedback, comment and tips to improve much appreciated.

The goal of developing PrivacyTrail was to allow organisations of all sizes to assess their digital privacy, especially those who can't afford expensive threat intelligence tooling or haven't got the time to do it themselves.


Had no idea about this, thanks


I have actually written a tool to help organisations understand their own digital exposure/privacy. It is very much an MVP but check it out, it should help you with some of your concerns - https://www.privacytrail.com

I think DDR0 makes a strong point, a determined malicious actor will always find a domain you haven't considered so defensively registering dozens of domains is usually only an action taken by large banks or significant brands. Unless you are likely to be impersonated or have a duty of care similar to that of a bank then a single/small number of domains is probably sufficient, especially for a startup. Buying additional domains can be done as you get larger and the threat of impersonation increases.

However, that doesn't mean you shouldn't monitor domain purchases that are similar to your own. Blacklisting domains that you believe have been purchased for nefarious goals can prevent your own employees from being duped in convincing phishing attacks and it is always good to occasionally remind customers/third parties of the domains you operate from.

Anyway, I could waffle about this for ages - there is more info on the above link and you can try your own domain out!


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: