What are the opening hours?

Monday 9am - 5pm Tuesday 9am - 5pm Wednesday 9am - 5pm Thursday 9am - 5pm Friday 9am - 5pm

Cambridge Micrologic

Home
United Kingdom
Cambridge
Cambridge Micrologic

Machine intelligence that predicts hardware-failures.

05/01/2022

All electro-mechanical systems exhibit a non-linear behaviour before they stop working, or need servicing. The trick it so catch them before they stop working. You need to spot issues before they become urgent. Our AI machine intelligence can automatically spot strange behaviours. It is the extra set of eyes on your system 24/7. AI is more than just alert level detection, it gives you an insight as to what is going to happen and when.

21/10/2021

Cambridge Micrologic deliveres AI at the point of application. Systems that monitor machinery and reports in real-time on asset performance. This is predictive maintenance. Report are straight to your mobile device as text messages. It is the fastest condition monitoring with reporting. It is the solution for critical assets.

14/10/2021

There is a new standard in town for radio communications equipment. EN300328 has been updated for devices that use ISM band. This includes Bluetooth, Wi-Fi and Zigbee. The change took effect on 6 August 2021.

The changes cover receiver blocking requirements, spurious radio emission limits and a revision of equipment classifications.

Receiver blocking happens when strong radio emissions from unwanted sources intrude on communication channels. This causes a reduced performance in the receiver circuitry.

07/10/2021

Knowing what the temperature is of your asset, and how it is going to change, is vital to its reliability. AI predicts trends, so you know when action is needed.

Temperature sensors come in many types. The right sensor is needed for your application. Then custom AI monitors and reports via SMS straight to your mobile device. Machine intelligence, at the point of application, analysis then reports in real time. It is the extra resource to ensure your asset keeps going.

21/09/2021

The use of industrial digital sensors simplifies integrating AI into an existing machine. Often these are already fitted. Their signals can be eavesdropped, then analysed.

Tapping into existing sensors enables fast deployment of data analysis of thresholds being exceeded. Analysing all sensors’ data streams in real time, then reporting via 4G SMS means that no extra programming is required. It makes the work of introducing predictive maintenance as simple as possible.

14/09/2021

Cable are the root cause of most problems for product reliability and approvals. Cables are antenna for electrical noise getting into your system. It can wreak havoc with the power supply, and corrupt data signal integrity. For approvals, electronic products must not emit radio waves above a given level. They must also be immune to a certain level. Screen cables and metal boxes are not always the answer. The topology of the design is key to success.

#

08/09/2021

FemtoML machine learning is for progressing customer needs in smarter devices. It does this without adding large computer code. This is done with small code. For example, it can monitor and advise on power consumption.

FemtoML solves the problem of a processor having to do a lot of number crunching to predict an outcome. It ‘measures’ its environment, then evaluates that result. From this, it makes predictions.

01/09/2021

Why would you want an analogue computer for AI? It sounds pre-1970’s The answer is that they can do large complex measurements without burderning the microprocessor with large code routines. This enables you to use small memory devices and use much less power.

For analysing frequency bands, this is a huge advantage. In the digital world to analayse the audio band from 500Hz to 20KHz, with a resolution of 20Hz, requires 2900 calculations to be held in memory. For many microcontroller chips, the memory space for this is not an option. 46k of RAM is required. Many small memory devices have much less than this for RAM. It also enables frequency band analysis to be added to exisiting systems without a large overhead on processor requirements.

10/08/2021

Artificial intelligence computing is the means to enabling a job to be done. That job might be to keep a motor going in plant machine. Nobody wants their machine to stop working.

To keep a motor going, it will need maintenance. AI will monitor motor parameters, then predict what is likely to happen and when. This is a lot of number crunching. For a critical system, it must fast and live. An analogue computer will do calculations in an instant.

There are companies developing analogue computers for AI. They include IBM Research, Mythic and Cypress Semiconductor. The job to be done with these devices is to apply them for condition monitoring to keep a motor going.

Analogue computing is back.

03/08/2021

Analogue-digital hybrid AI computing is a growing technology. Its application is for high-speed condition monitoring. Uses include induction motors.

For vibration monitoring, there are over twenty parameters that can be used to profile a performance. Some of these can be done with analogue electronics. Common ones are frequency analysis, crest factor and root mean square (RMS).

Developments using low-power microcontrollers gives you four to six analogue inputs. Analogue electronics gives a good real-time indication of the above parameters, and without using any coding. The microcontroller is then free process the results to deduce a performance trend as machine learning.

28/07/2021

Analogue AI-computing is different to its digital counterpart. You are looking for the resulting output. How it works out calculations is where you need to change your mindset. Analogue computers give a different type of result.

For example, if you were analysing a song for high pitch notes, analogue will instantly give you a band of frequencies that are present. Whereas a digital system will give you specific frequencies, after a lot of calculations. If you want precision, then digital computing is your solution. If you want a trend for what what is happening and to know that there are high pitches, analogue will deliver.

20/07/2021

Analogue-digital hybrid electronics for artificial intelligence is suitable for many applications. For sensor signals that are continuous. Analogue processing is instantaneous, low-power and has minimal electromagnetic emissions. It is a perfect solution.

The first pitfall to avoid is trying to replicate digital processing with analogue electronics. That will not work, so do not do that. AI is about estimating outcomes. Analogue processing does that.

For an induction motor with an imbalance problem, it may be spinning at 300-500rpm. This could be your washing machine. Detecting frequencies that show the issue can be done with analogue electronics. No digital electronics are required until the final stage of processing.

Address

Cambridge
CB43AZ

Opening Hours

Monday	9am - 5pm
Tuesday	9am - 5pm
Wednesday	9am - 5pm
Thursday	9am - 5pm
Friday	9am - 5pm

Website

http://www.micrologic.ai/

Alerts

Be the first to know and let us send you an email when Cambridge Micrologic posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Contact The Business

Send a message to Cambridge Micrologic: