S5E6 - Connected Care: The Future of Patient Monitoring

In this episode of AXREM Insights, Sally Edgington and Melanie Johnson sit down with Chris Lawman (GE Healthcare) and David Britton (Mindray) to explore the work of AXREM’s Patient Monitoring Special Focus Group. Chris and David bring decades of experience in the medtech field, discussing their passion for patient care and how collaboration across the industry is driving positive change. A central theme is the group's recently launched manifesto, built around the “four Cs” – Consistency, Connectivity, Care Innovation, and Cash Savings – which aim to improve NHS workflows, leverage existing equipment more effectively, and ultimately deliver better patient outcomes.

The conversation highlights the growing importance of connectivity and interoperability within hospitals, and how industry input is vital to influencing government policy and NHS direction. They also touch on the future of patient monitoring, including wearable technologies, virtual wards, and the role AI will play in early intervention and personalised care. Throughout the episode, Chris and David emphasise the value of AXREM as a voice for the medtech industry – creating a safe, collaborative space where members can raise shared challenges, influence policy, and shape the future of UK healthcare together.

Rad Magazine sponsor of AXREM's UKIO Drinks Reception and Leading Publication in the Medical Imaging and Oncology Space

Thanks for listening to this week's episode

To find out more about AXREM check out our website HERE
If you are interested in joining AXREM as a member CLICK HERE
To contact us CLICK HERE

And join us next time for more insights from industry.

Transcript

This Transcript is just a guide and is auto generated, there maybe spelling errors but this is just an aid to listening in.

[00:00.000 --> 00:03.000]  Hello, I'm Melanie Johnson and I'm here with Sally Edgington.
[00:03.000 --> 00:06.000]  Today we have the pleasure to be speaking to both Chris Lawman,
[00:06.000 --> 00:10.000]  Patient Monitoring and Accessories Modality Manager at GE Healthcare
[00:10.000 --> 00:14.000]  and also AXREM's Patient Monitoring Special Focus Group Convener
[00:14.000 --> 00:18.000]  and also David Britton, Patient Monitoring and Life Support Marketing Manager
[00:18.000 --> 00:20.000]  at Modality Manager at GE Healthcare.
[00:20.000 --> 00:23.000]  Today we have the pleasure to be speaking to both Chris Lawman,
[00:23.000 --> 00:27.000]  Patient Monitoring and Accessories Modality Manager at GE Healthcare
[00:27.000 --> 00:31.000]  and David Britton, Patient Monitoring and Life Support Marketing Manager at MindRay
[00:31.000 --> 00:34.000]  and AXREM's Patient Monitoring Special Focus Group Vice-Convener.
[00:34.000 --> 00:38.000]  So welcome both Chris and David and thank you for being on the show today.
[00:38.000 --> 00:43.000]  Now let's get started by handing over to you to tell us a little bit about yourself and what's your story.
[00:43.000 --> 00:46.000]  Thank you Mel. So my name is Chris Lawman.
[00:46.000 --> 00:54.000]  I am the Patient Monitoring Manager for GE Healthcare and I cover Northern Europe.
[00:54.000 --> 00:59.000]  I've only recently just been appointed the Northern European role.
[00:59.000 --> 01:04.000]  I've been with GE Healthcare for coming up to 28 years now,
[01:04.000 --> 01:08.000]  where I started off with a company that was called Market Heliger
[01:08.000 --> 01:13.000]  and we were acquired by GE back in 1999, 2000.
[01:13.000 --> 01:18.000]  And as you can imagine over those 20 odd years I've had quite a number of roles,
[01:19.000 --> 01:24.000]  but I spent the large part of my time in GE Healthcare in service,
[01:24.000 --> 01:28.000]  where I used to obviously repair equipment,
[01:28.000 --> 01:31.000]  but then I went into more the IT side of things.
[01:31.000 --> 01:40.000]  So I used to help train and design networks for our equipment that go into the hospitals.
[01:40.000 --> 01:45.000]  And then it's been the last 10 years where I actually moved over to the sales side of the business,
[01:45.000 --> 01:50.000]  working on patient monitoring, focusing on just patient monitoring.
[01:52.000 --> 01:54.000]  So you're obviously enjoying it.
[01:54.000 --> 02:00.000]  I think that's the thing that really helps. I'm really passionate about it.
[02:00.000 --> 02:03.000]  Many of those years David and I worked together.
[02:03.000 --> 02:12.000]  We collaborated a lot on certain products and I think we both had a similar approach to the business.
[02:12.000 --> 02:17.000]  We had a passion, we had a desire and I think that's what keeps us going.
[02:20.000 --> 02:24.000]  And David, obviously we were fortunate to have you on our show.
[02:24.000 --> 02:26.000]  I think it was series one Meet the Team.
[02:26.000 --> 02:29.000]  So obviously it was nice to see you then and it's great to have you back on.
[02:29.000 --> 02:31.000]  So thank you and welcome.
[02:31.000 --> 02:35.000]  So again, if you just want to give us a little bit of a highlight about obviously your background
[02:35.000 --> 02:39.000]  and I suppose I'm going to ask you just an extra question actually.
[02:39.000 --> 02:42.000]  What was the key moment in your career? What was that?
[02:42.000 --> 02:45.000]  What was kind of that highlight moment that really stood out?
[02:45.000 --> 02:49.000]  I mean I spent all that time with Chris. I spent 20 years working with Chris.
[02:49.000 --> 02:51.000]  That's got to be the highlight.
[02:51.000 --> 02:53.000]  I prompted the question though.
[02:57.000 --> 02:59.000]  I will come back to that one.
[02:59.000 --> 03:05.000]  But honestly I've enjoyed my time at GE. I had 28 years at GE so that was good fun.
[03:05.000 --> 03:13.000]  I'm the Modality Manager over at Mindray where we supply solutions to the NHS and private sector.
[03:13.000 --> 03:19.000]  And like Chris, you love it. Once you fall into healthcare, it really is something special.
[03:19.000 --> 03:22.000]  Yes, there are challenges and there are challenges every day.
[03:22.000 --> 03:28.000]  But honestly it is delivering for your customers, delivering ultimately for the patients.
[03:28.000 --> 03:30.000]  So it just gives you an absolute buzz.
[03:31.000 --> 03:35.000]  One career highlight. I love what I do now and I love what I've done in the past.
[03:35.000 --> 03:40.000]  But in between there was a hybrid theatre environment which I absolutely adored.
[03:40.000 --> 03:43.000]  So I had a really good time in the imaging sector.
[03:43.000 --> 03:48.000]  So for me that was developing a totally new area of the market and that was great fun.
[03:48.000 --> 03:50.000]  And that was great fun.
[03:50.000 --> 03:58.000]  Yeah and I think our industry is about the people and whether that's the staff or the patients.
[03:58.000 --> 04:01.000]  So it is a people industry. So yeah, I absolutely get that.
[04:01.000 --> 04:04.000]  And I can see there's a real passion from our members.
[04:04.000 --> 04:12.000]  Very ethical people that want to do an ethical job that can indirectly or directly make a real difference to people's lives.
[04:12.000 --> 04:19.000]  So I think for our listeners it would be really good for them to get a bit of an understanding of the special focus groups,
[04:19.000 --> 04:27.000]  the patient monitoring special focus group and also the manifesto which we launched back in June 2024 at EBME Expo.
[04:27.000 --> 04:32.000]  And I know that I remember David, us sitting there, we came up with the four C's.
[04:32.000 --> 04:38.000]  Consistency, connectivity, care, innovation and cash savings.
[04:38.000 --> 04:45.000]  So I wondered actually whether Dave, you want to talk a little bit about the manifesto because I know you were quite instrumental in helping write that.
[04:45.000 --> 04:50.000]  And then Chris, if you want to talk a bit more about the special focus group and what the kind of priorities are.
[04:50.000 --> 04:52.000]  So should we go to Dave first?
[04:52.000 --> 04:54.000]  Yeah, no, thank you for that.
[04:54.000 --> 05:01.000]  And it was important because we had to give a sense of direction for the group and to say who we were.
[05:01.000 --> 05:05.000]  So all those stakeholders externally and internally knew what we were about.
[05:05.000 --> 05:11.000]  It's interesting as time goes by, some take priority over others.
[05:11.000 --> 05:20.000]  And I think within the industry at the moment, connectivity is certainly one of those key elements that we really focus on.
[05:20.000 --> 05:23.000]  And how we release that innovation into the market.
[05:23.000 --> 05:29.000]  All our members have wonderful, wonderful solutions.
[05:29.000 --> 05:39.000]  And we need to work with the NHS and the influencers there to release the potential of what we can offer.
[05:39.000 --> 05:47.000]  Because we always tend to sell a dream and then it's delivering on that dream and releasing the potential to our customers.
[05:47.000 --> 05:51.000]  So honestly, the manifesto still stands.
[05:51.000 --> 05:53.000]  It is still very relevant.
[05:53.000 --> 05:59.000]  And I'm delighted that Chris has taken on the mantle for this and to push it forward.
[05:59.000 --> 06:03.000]  Well, it's great that you laid the foundation, David, with the team, obviously.
[06:03.000 --> 06:09.000]  And I couldn't have put it better in terms of the four Cs in what you're focusing on.
[06:09.000 --> 06:16.000]  I've already said that I worked in connectivity in my role for quite some time.
[06:16.000 --> 06:22.000]  So the connectivity element has been, you know, a huge part of my life and a huge passion for me.
[06:22.000 --> 06:36.000]  And, you know, it showed that with the voice that Axrem has, that we had the ability to write the document that we sent to the government about connectivity, about going back to basics.
[06:36.000 --> 06:43.000]  You know, obviously there's the mantra with the NHS right now of big data and artificial intelligence.
[06:43.000 --> 06:47.000]  But, you know, we need to get that data to where it needs to be.
[06:47.000 --> 06:56.000]  And, you know, the NHS needs to tie and get back to basics and provide us with those networks in theatres and in the locations where they're needed.
[06:56.000 --> 07:07.000]  And, you know, without Axrem, we wouldn't have been able to get that viewpoint across on the basis that it comes from all the vendors that have a vested interest.
[07:07.000 --> 07:12.000]  Absolutely. And just to highlight just a couple of things that are some of the shocking figures actually.
[07:12.000 --> 07:17.000]  But first of all, in terms of the four Cs of consistency, it's about talking as one NHS.
[07:17.000 --> 07:21.000]  And I sound like a broken record these days. We talk about one NHS.
[07:21.000 --> 07:26.000]  We've got all the individual trusts and it adds a lot of complexity because they're not quite aligned.
[07:26.000 --> 07:30.000]  And obviously that's what as an industry we would like to see one NHS.
[07:30.000 --> 07:41.000]  In terms of care innovation, these figures shock me that the UK has one of the lowest numbers of hospital beds per 100,000 population compared to similar size country.
[07:41.000 --> 07:47.000]  So for the UK, it's 2.43 hospital beds per 100,000 people.
[07:47.000 --> 07:50.000]  In France, it's 5.7, so over double.
[07:50.000 --> 07:55.000]  And in Germany, 7.8, so triple per 100,000.
[07:55.000 --> 08:01.000]  So not having enough beds can lead to high bed occupancy rates and workflow issues.
[08:01.000 --> 08:11.000]  And then in terms of the connectivity piece, investing in connectivity is vital to transfer meaningful data to clinicians to improve the patient outcomes.
[08:11.000 --> 08:22.000]  And I would urge people to read the manifesto because there are some clear kind of calls to action on the back page around streamlined processes and procedures,
[08:22.000 --> 08:28.000]  internal hospital stakeholders working towards a common goal with clearly defined roles and responsibilities.
[08:28.000 --> 08:31.000]  We want trusts to unlock the potential of their current equipment.
[08:31.000 --> 08:35.000]  A lot of the equipment can do a lot more than what it's being used for.
[08:35.000 --> 08:47.000]  So while they've made that investment, use it to its maximum potential and be able to have more data around how your patients, their vital signs are performing.
[08:47.000 --> 09:01.000]  And then cash savings, increased efficiency of workflows will support improved outcomes, early mobilization of patients and cut waiting lists, which is one of the government's key things is to cut waiting lists.
[09:01.000 --> 09:11.000]  Honestly, and you look at the elective surgery data still, you know, it's up at 7 million folk waiting for their operations, which is huge.
[09:12.000 --> 09:16.000]  And I know that that's still the backlog from COVID, but that needs to be cleared.
[09:16.000 --> 09:23.000]  And you're spot on when you talk about early mobilization of patients, because that's what's going to help here.
[09:23.000 --> 09:32.000]  Longer term, and I know this isn't new, but we speak about the aging population within the UK, and that's still relevant.
[09:32.000 --> 09:36.000]  So the situation is only going to sort of worsen.
[09:36.000 --> 09:39.000]  So how do we look after those people? How do we keep them safe?
[09:39.000 --> 09:44.000]  And it's that joined up thinking between hospital and the social care element.
[09:44.000 --> 09:46.000]  And how do we join those two together?
[09:46.000 --> 09:48.000]  There's there's so much there.
[09:48.000 --> 09:53.000]  I think we also I think we all love the NHS and we want it to work.
[09:53.000 --> 09:57.000]  But there are so many elements that need improvement, you know.
[09:57.000 --> 10:03.000]  And gosh, we hope that we're working as part of that to bring better solutions.
[10:04.000 --> 10:07.000]  There's a frustration, isn't there, because it touches all of us.
[10:07.000 --> 10:13.000]  You know, I know that we've all been touched personally by the requirements of what the NHS delivers.
[10:13.000 --> 10:20.000]  So you've kind of got a double vested interest in it in terms of what it can deliver for your family and your friends.
[10:20.000 --> 10:29.000]  But actually how professionally you can assist, you know, those caregivers and ultimately those patients.
[10:29.000 --> 10:33.000]  And again, that comes back to the passion and why we do our job.
[10:33.000 --> 10:34.000]  Absolutely.
[10:38.000 --> 10:43.000]  I know we've obviously had conversations and obviously looking further ahead in horizon scanning.
[10:43.000 --> 10:46.000]  I want to talk about what's next for the special focus group.
[10:46.000 --> 10:57.000]  And like I say, I know our conversations have been very based around trying to encourage our members to engage a little bit more on any challenges and key topics, which we can obviously look into further.
[10:57.000 --> 11:04.000]  So if you can just kind of go into that a little bit more and obviously why it's important as well that we do really encourage our members to speak out.
[11:04.000 --> 11:08.000]  I've kind of already touched on the point that, you know, Axrem gives us a voice.
[11:08.000 --> 11:15.000]  You know, it gives us that ability to get a collaborative approach.
[11:15.000 --> 11:17.000]  You know, we've all got a vested interest.
[11:17.000 --> 11:27.000]  We're all in the same boat in terms of some of the challenges that are presented to us from working with, you know, the likes of the NHS.
[11:27.000 --> 11:33.000]  So, you know, one thing that we really need to drive is to sort of further work on that collaboration.
[11:33.000 --> 11:36.000]  What we've got to remember, it's an ever moving industry.
[11:36.000 --> 11:45.000]  You know, there's things changing all the time, whether that be technology, whether that be the government, whether that be funding.
[11:45.000 --> 11:51.000]  You know, there's all these different factors that impact us within industry.
[11:51.000 --> 12:03.000]  So it's very difficult to try and put a subject in terms of what we're going to focus on next because we don't know necessarily what's going to come along.
[12:03.000 --> 12:13.000]  There's some obvious ones that we've been working on in the past, like the challenges around, again, interoperability, the technology, the famous DTAC forms that we have to fill in.
[12:13.000 --> 12:19.000]  And, you know, these kind of elements that cause us all the same kind of challenge.
[12:19.000 --> 12:26.000]  But to be able to pinpoint something that we're going to focus on next month or the month after is really, really difficult.
[12:26.000 --> 12:35.000]  And this is where we really need the members of the special focus group to bring those subjects to our attention, to bring them to your attention.
[12:35.000 --> 12:48.000]  You know, if I think that it's obviously competitive or you've got some competitive challenges because of, you know, we're all businesses working in the same sort of area,
[12:48.000 --> 12:57.000]  then, yeah, come out to you guys, reach out, ask the questions before bringing it to the special focus group and talking about it in an open forum.
[12:57.000 --> 13:02.000]  But, yeah, you know, we're here to listen. We're here to work on it if we if we can.
[13:02.000 --> 13:09.000]  And, you know, we're waiting to hear what the members have got to say.
[13:09.000 --> 13:12.000]  And I think it's really nice as well because we are quite a small team.
[13:12.000 --> 13:19.000]  We can be very agile and I like to think that we are building that real sense of community as well.
[13:19.000 --> 13:28.000]  So hopefully people do feel comfortable to speak and have those conversations, whether it's with us personally as in Axrem or within a special focus group.
[13:28.000 --> 13:40.000]  So, like I say, any point anyone can reach out to us, we're happy to have those conversations and raise those because I'm sure if one person's having a particular challenge,
[13:40.000 --> 13:45.000]  somebody else will be. So and that's what we're kind of noticing more and more.
[13:45.000 --> 13:52.000]  I agree entirely with that. It is giving everybody the ability to to feel comfortable to speak.
[13:52.000 --> 14:02.000]  And it's honestly that you've got to spot on when when you say that if you bring up a topic, it will affect more than just your good self.
[14:02.000 --> 14:08.000]  So it's OK to bring it up because there is something I should go back to DTAC again.
[14:08.000 --> 14:10.000]  It was always a fear. Should we bring it up? Should we?
[14:10.000 --> 14:13.000]  You know, everybody had challenges with it, but guess what?
[14:13.000 --> 14:17.000]  It's an industry challenge. So if it's an industry challenge, let's get it sorted.
[14:17.000 --> 14:20.000]  And I love what Axrem has done and given us that voice.
[14:20.000 --> 14:26.000]  So, yeah, it's a safe space, David. It's a safe space to come and open up and talk about it.
[14:26.000 --> 14:30.000]  And if we feel we can't do anything about it, we'd probably talk around that as well.
[14:30.000 --> 14:32.000]  You know, it's gone.
[14:32.000 --> 14:46.000]  And I think as well, when we're horizon scanning to kind of what's next for the SFG, what's your kind of thoughts on one of the main pledges of the Department of Health and Social Care is hospital to home, so hospital to community.
[14:46.000 --> 14:56.000]  So what's your kind of thoughts on wearable patient monitoring devices, virtual wards, because I think that's definitely going to be something that we will work on moving forward.
[14:56.000 --> 14:59.000]  And I think there'll be a lot more emphasis on that as well.
[14:59.000 --> 15:16.000]  I think it's a really important element because we have to adapt to the needs of the industry as well or the hospitals and that moving people to home, caring for people at home so that you're stopping people actually moving back into hospitals.
[15:17.000 --> 15:20.000]  So if they have, you know, stopping that need for acute care, because guess what?
[15:20.000 --> 15:26.000]  You made an early intervention or making sure that people are early mobilised so they can get home.
[15:26.000 --> 15:28.000]  That's really important. It will help with the waiting lists.
[15:28.000 --> 15:38.000]  It will help us all get the care that we need for ourselves, for our loved ones and for the public in general.
[15:38.000 --> 15:43.000]  Yeah, I was going to say it's the hot topic right now.
[15:43.000 --> 15:53.000]  I had the benefit this week of attending the Association of Surgeons, sorry, Great Britain and Ireland, and it was a massive factor for the surgeons.
[15:53.000 --> 16:02.000]  Obviously, they have to look after their patients during their procedure, but obviously they've got a vested interest in their recovery.
[16:02.000 --> 16:07.000]  And, you know, their recovery period is hopefully a short time in the hospital environment.
[16:07.000 --> 16:19.000]  And then a lot of it is obviously once they've gone home and, you know, so it's huge in terms of what can be done from a home health perspective.
[16:19.000 --> 16:23.000]  I sometimes sit back and think, how is it handled?
[16:23.000 --> 16:35.000]  How, if we've got a lot of equipment, if we've got a lot of patients in the home, in the field that are, you know, utilising equipment, how is it triaged?
[16:35.000 --> 16:44.000]  How is it handled from a, you know, I don't know, which care givers look after them and who sends the ambulances?
[16:44.000 --> 16:47.000]  And does it fall into primary? Does it fall into secondary care?
[16:47.000 --> 16:50.000]  You know, a lot of challenges out there.
[16:50.000 --> 17:00.000]  Yeah, and we were recently talking about that at our MedTech conference and there's that divide between primary and primary care and social care.
[17:01.000 --> 17:04.000]  And obviously they have separate budgets, so they've got a ring fence.
[17:04.000 --> 17:08.000]  But treating people at home is always better for the patient.
[17:08.000 --> 17:17.000]  You know, I've been in hospital a lot in my life and I would much rather be at home recuperating than in a hospital bed where the wards are noisy.
[17:17.000 --> 17:21.000]  There's lots of beeping, probably your machine's beeping.
[17:21.000 --> 17:27.000]  Our machine's beeping, very cool.
[17:28.000 --> 17:35.000]  But yeah, and also it's better for the families in terms of having to travel back and forwards and managing their work and other commitments.
[17:35.000 --> 17:43.000]  But it's also better for the taxpayer, you know, and actually that is, you know, why this is such a focus for the government.
[17:43.000 --> 17:56.000]  And I think that you can see that this government wants patients to have more choice and, you know, and decide more about their care.
[17:56.000 --> 18:06.000]  And absolutely, that's the right approach because we are very patient focused and industry as in patient focus is what the government are, which is great.
[18:06.000 --> 18:10.000]  So I think that a lot of our messaging is quite aligned to the government.
[18:10.000 --> 18:19.000]  And I think that next month in June, we will be seeing the 10 year plan and the spending review.
[18:19.000 --> 18:22.000]  So I think it'll be quite interesting to see what those focuses are.
[18:22.000 --> 18:30.000]  And obviously, as Axel and we will be scrutinising it with a very, very close eye and seeing what is relevant to our members.
[18:30.000 --> 18:32.000]  And that will help guide you, Chris.
[18:32.000 --> 18:39.000]  I think when you develop this group, you know, we'll see some key focuses come out of that and work that we'll probably want to do.
[18:39.000 --> 18:52.000]  Absolutely. I think, well, we've got some sort of immediate sort of challenges coming up in terms of understanding how the NHS is going to be set up with NHS England now with all of those changes.
[18:52.000 --> 18:55.000]  And we're kind of yet to see how that's going to pan out.
[18:55.000 --> 19:08.000]  But you're absolutely right. We have to look at the bigger picture here and see what, well, how that's going to impact us, but actually how we can adopt those changes and help our customers at the same time.
[19:08.000 --> 19:16.000]  Yeah, and I think in terms of the abolishment of NHS England and integrate it into the Department of Health and Social Care, obviously that was a previous model.
[19:16.000 --> 19:18.000]  This model may look different.
[19:18.000 --> 19:22.000]  And I think it's important to understand and it's got to be executed really well.
[19:22.000 --> 19:27.000]  But if it is, it will take out another layer of complexity.
[19:27.000 --> 19:35.000]  So I'm hoping that this plays into our asks and our foresees, you know, and that we are more consistent.
[19:35.000 --> 19:37.000]  So I hope that that's what happens.
[19:37.000 --> 19:42.000]  I think it's going to be a tricky few years when the government made the announcement.
[19:42.000 --> 19:44.000]  They said it was going to be two years.
[19:44.000 --> 19:46.000]  I think that's quite ambitious.
[19:46.000 --> 19:56.000]  But equally, you have to feel for a lot of people that work within NHS England in the function there that, you know, they don't know whether they've got a job in two years time.
[19:56.000 --> 20:06.000]  So I think for us that, you know, we've been working really closely with a lot of NHS England colleagues who have been fantastic and we've had really good engagement.
[20:06.000 --> 20:11.000]  And I hope those people are retained within this process.
[20:11.000 --> 20:15.000]  So I think the uncertainty could affect our engagement.
[20:15.000 --> 20:18.000]  And that's really important as a trade association that we don't lose that.
[20:18.000 --> 20:28.000]  So that'll be my job to make sure that we have that consistency with our communications with the Department of Health and NHS England whilst it exists.
[20:28.000 --> 20:31.000]  Good, good, yeah.
[20:31.000 --> 20:42.000]  So I'd like to now ask you sort of more broader, what you think the benefits of being involved with the special focus group and acts more widely.
[20:42.000 --> 20:45.000]  So if we start with you, Chris, first.
[20:45.000 --> 20:48.000]  Yeah, I think I've already sort of mentioned a part of it.
[20:48.000 --> 20:54.000]  The fact that we've got the whole sort of industry represented within the special focus group.
[20:54.000 --> 20:56.000]  You know, I couldn't tell you the number.
[20:56.000 --> 21:03.000]  How many how many businesses are represented in the special focus group right now, Sally, within patient monitoring?
[21:03.000 --> 21:05.000]  Was it 12, 14 of us?
[21:05.000 --> 21:06.000]  Something like that.
[21:06.000 --> 21:07.000]  I think so.
[21:07.000 --> 21:14.000]  You carry on talking and I'll tell you how many it is in just a second because you've put me on the spot now.
[21:14.000 --> 21:21.000]  But it goes to show that, you know, there's so many people that have come to the special focus group that they see a value in it.
[21:21.000 --> 21:28.000]  And the fact that we've got that collaborative approach, as I've already said, you know, it's a flexible moving industry.
[21:28.000 --> 21:30.000]  It's ever changing.
[21:30.000 --> 21:39.000]  And therefore, you know, the different partners that we have within the business really want to understand what's going on.
[21:39.000 --> 21:44.000]  I think it gives you in a strange way that little bit of peace of mind.
[21:44.000 --> 21:52.000]  You know, you're not a lone wolf when you're looking at these certain challenges.
[21:52.000 --> 21:54.000]  You know, I think David alluded to it earlier.
[21:54.000 --> 21:56.000]  You know, do we bring this subject up?
[21:56.000 --> 21:57.000]  Have we got to talk about this?
[21:57.000 --> 22:01.000]  But actually, you know, it comes back to sharing is caring, doesn't it?
[22:02.000 --> 22:13.000]  You know, it's nice to say that, you know, I could sit in the corner of a room, have a chat with somebody and actually they understand exactly what we're going through.
[22:13.000 --> 22:20.000]  And then you can share those anxieties and whatever way it impacts you, you know.
[22:20.000 --> 22:30.000]  But I've already mentioned it, the voice, the fact that, you know, you've got these great contacts with various members of the government, various different bodies.
[22:30.000 --> 22:33.000]  That allow us to get our point across.
[22:33.000 --> 22:47.000]  You know, it was fantastic last year to come to the Houses of Parliament to one of the launches there and, you know, actually get time to sit and speak to some of the MPs about, you know, some of our challenges.
[22:47.000 --> 22:53.000]  So, yeah, no, it's just that voice and that collaboration, I think, that helps us.
[22:53.000 --> 22:58.000]  You can continue the quiz now and ask me the question.
[22:58.000 --> 23:01.000]  Your starter for one, Sally.
[23:01.000 --> 23:05.000]  13 companies, we have 13 companies in there.
[23:05.000 --> 23:07.000]  You did, you were not too far off.
[23:07.000 --> 23:15.000]  And what I would also say as well is, I think that the group will, it's obviously still quite a young group within Axrem.
[23:15.000 --> 23:19.000]  And so I think we will see much more.
[23:19.000 --> 23:22.000]  Definitely not you guys.
[23:22.000 --> 23:29.000]  Don't worry, we'll ask you back again to do a podcast.
[23:29.000 --> 23:31.000]  You can get your own back.
[23:31.000 --> 23:39.000]  But no, I think as well, it's really good that we have, for instance, NHS supply chain attend a lot of the meetings and we're really trying to be quite collaborative.
[23:39.000 --> 23:41.000]  And that is key in all of Axrem's work.
[23:41.000 --> 23:50.000]  And I think when we've got a younger group within Axrem, that's going to be key to kind of pushing the visibility of the group, the work that you guys are doing.
[23:50.000 --> 23:53.000]  Obviously, we're working quite closely as well with eBME.
[23:53.000 --> 23:58.000]  I joined there in all, I think it was called an eBME Leaders Forum.
[23:58.000 --> 24:02.000]  And I've got another meeting coming up, kind of representing you guys within that forum.
[24:02.000 --> 24:04.000]  That was really interesting.
[24:04.000 --> 24:06.000]  And I took quite a lot of way from that.
[24:06.000 --> 24:11.000]  So, yeah, I think that there's a lot of collaboration to do going forward.
[24:11.000 --> 24:17.000]  So over to you, Dave, what do you think are the kind of benefits of membership of Axrem?
[24:17.000 --> 24:20.000]  I mean, you've hit on a lot of them already.
[24:20.000 --> 24:24.000]  But for me, it's sort of affecting meaningful change.
[24:24.000 --> 24:34.000]  So as an organisation, it's able to represent the industry with all those key stakeholders, the people that are making decisions.
[24:34.000 --> 24:42.000]  If they're making decisions with only one side of the sort of discussion, then not everything is going to get implemented.
[24:42.000 --> 24:46.000]  So it feels like a much healthier environment.
[24:46.000 --> 24:56.000]  If the people that are supplying a lot of the services to the NHS are able to be part of that discussion and affect how the NHS is developed,
[24:56.000 --> 25:04.000]  you know, see what's realistic, see what isn't realistic, where we can support and within still within that competitive environment.
[25:04.000 --> 25:09.000]  So nothing's lost from the competitive edge, but you're getting that industry voice.
[25:09.000 --> 25:11.000]  And I think that's really important.
[25:12.000 --> 25:22.000]  And some of the stuff I've loved, like Chris, the Houses of Parliament and seeing the government were interacting with Axrem,
[25:22.000 --> 25:30.000]  but also at a local level, if it wasn't for your initiative, saying, hey, why don't you go out and get your local MPs involved?
[25:30.000 --> 25:33.000]  Honestly, on a local level, that's worked really well.
[25:33.000 --> 25:35.000]  And it's yeah, that's great.
[25:35.000 --> 25:39.000]  So that engagement, meaningful change works for me.
[25:39.000 --> 25:40.000]  Great, thank you.
[25:42.000 --> 25:48.000]  Right, so I'm just going to kind of lighten the mood slightly now and go on to my quirky question, as I always do.
[25:48.000 --> 25:54.000]  OK, so if you had a time machine, would you choose to go back in the past or go to the future and why?
[25:54.000 --> 25:56.000]  And David, I'll go to you first.
[25:59.000 --> 26:00.000]  Here we go. Here we go.
[26:00.000 --> 26:03.000]  I do like music and I do like live music.
[26:03.000 --> 26:10.000]  And I think I may have said in my earlier podcast that I have an earworm of the moment, which is a band called Pete and Diesel.
[26:10.000 --> 26:13.000]  So I went to see them up in a festival in Scotland recently.
[26:13.000 --> 26:20.000]  However, however, if I could have a time machine, I'd love to go back to Woodstock, to be quite frank.
[26:20.000 --> 26:26.000]  I think I'd love to see Woodstock and be there at Woodstock and maybe with a clearer mind than most people there.
[26:26.000 --> 26:28.000]  But, you know, we'll see.
[26:28.000 --> 26:36.000]  I swear, Dave, that you must be getting a proportion of royalties for this group, because you mentioned that last time too.
[26:39.000 --> 26:41.000]  It's true, can't help it.
[26:43.000 --> 26:45.000]  They're family friends really, aren't they?
[26:45.000 --> 26:47.000]  It's absolutely true.
[26:47.000 --> 26:49.000]  You get royalties on the tickets.
[26:49.000 --> 26:51.000]  Oh, brilliant, thank you.
[26:53.000 --> 26:55.000]  And Chris, over to you.
[26:55.000 --> 27:04.000]  Mine was going back into the past, but mine's perhaps a little bit boring because I didn't come into this industry until I was 27 years old.
[27:04.000 --> 27:10.000]  And the more I've learned through working in this industry, I really wish that I'd have been able to get into it a little bit earlier.
[27:10.000 --> 27:15.000]  When I left school, a spotty teenager that didn't really know what direction to go in.
[27:15.000 --> 27:19.000]  You know, I ended up working in the supermarkets, working in the pubs and all that sort of thing.
[27:19.000 --> 27:29.000]  I didn't really have any clear direction until, you know, I really kind of joined Marquette Helleger and then subsequently into GE.
[27:29.000 --> 27:37.000]  I wish that I'd had the clinical background or I'd done biomedical engineering or something like that from school.
[27:37.000 --> 27:39.000]  I really do.
[27:39.000 --> 27:45.000]  Plus, you know, going back to when I was sort of 16, 17, 18 years old, I'd be able to still hold a cricket bat.
[27:45.000 --> 27:54.000]  I'd still be able to play cricket instead of being, you know, half crippled in a chair at home, you know, in the corner on a Sunday afternoon watching the football.
[27:54.000 --> 27:57.000]  You know, I'd love to still be out there playing a bit of sport.
[27:57.000 --> 28:00.000]  But yeah, I think that's my one.
[28:00.000 --> 28:05.000]  You've got a much more wholesome outlook on life than I have, by the way.
[28:05.000 --> 28:08.000]  You make me look as if I'm in the shallow end.
[28:08.000 --> 28:11.000]  Not at all. Not at all, David. Not at all.
[28:11.000 --> 28:13.000]  Yeah, you know, it's funny, isn't it?
[28:13.000 --> 28:19.000]  It's just I guess if I'd have probably thought about it a little bit longer, I'd have come up with something maybe a little bit more lighthearted.
[28:19.000 --> 28:21.000]  But no, I do. I love my job.
[28:21.000 --> 28:29.000]  I love what I've done, you know, how much longer I've got left, who knows, before I, you know, hang up the laptop as it is these days.
[28:29.000 --> 28:33.000]  But no, I think I'll tell you what I saw the other day.
[28:33.000 --> 28:35.000]  Right. And it's changed my perspective.
[28:36.000 --> 28:42.000]  So at the Association of Surgeons, they have all these different sort of stands, obviously.
[28:42.000 --> 28:45.000]  And I love watching the old medical programs.
[28:45.000 --> 28:50.000]  And they had some of these areas where you could go and play where you actually practice doing sutures.
[28:50.000 --> 28:52.000]  And I actually had a go.
[28:52.000 --> 28:57.000]  And oh, my God, I've got this even more utmost respect for these surgeons.
[28:57.000 --> 29:03.000]  You know, when they are, you know, sewing things up that are millimeters thick, you know,
[29:03.000 --> 29:11.000]  I couldn't I couldn't even judge the depth, let alone trying to pick up the needle, then trying to turn it into the right position.
[29:11.000 --> 29:15.000]  It's insane how clever these guys are.
[29:15.000 --> 29:18.000]  Oh, I totally agree with you.
[29:18.000 --> 29:24.000]  I've had 17 maxillofacial operations and some of my surgery was very intricate.
[29:24.000 --> 29:32.000]  Thank God the surgeon didn't have the shakes because mine, you know, people generally wouldn't know I'd had all that surgery other than my lobstided smile.
[29:32.000 --> 29:34.000]  But I absolutely agree.
[29:34.000 --> 29:38.000]  And I actually resonate with you a lot, Chris, because I'm not from a clinical background.
[29:38.000 --> 29:42.000]  I left school not knowing what I wanted to do with absolutely no qualifications at all.
[29:42.000 --> 29:47.000]  And I got into this by accident 11 years ago into health care and trade associations.
[29:47.000 --> 29:49.000]  And I just got it. It just clicked with me.
[29:49.000 --> 29:52.000]  So I was a lot older than you were when you were 27.
[29:52.000 --> 29:55.000]  And it's the best thing that ever happened to me.
[29:55.000 --> 30:01.000]  And when you were saying you don't know how long you've got left, you know, obviously being only 28, I'm sure you've got lots of years left.
[30:01.000 --> 30:06.000]  I'm just trying to work out the numbers now actually.
[30:06.000 --> 30:10.000]  There you go. And I'm actually going to ask just a follow on question to that.
[30:10.000 --> 30:15.000]  How do you think AI will shape health care in 50 years time?
[30:15.000 --> 30:19.000]  What is what does 50 years look like with AI?
[30:19.000 --> 30:21.000]  What a question.
[30:21.000 --> 30:24.000]  I mean, you can see it already coming in now, can't you?
[30:25.000 --> 30:34.000]  Automatically detecting organs within ultrasound and, you know, products being able to communicate with each other and do decision making for us.
[30:34.000 --> 30:38.000]  This is the early stages in health care of AI.
[30:38.000 --> 30:40.000]  50 years time.
[30:40.000 --> 30:45.000]  It's quite scary to think about where technology is going to take us.
[30:45.000 --> 30:49.000]  It's a tough question, Sally.
[30:49.000 --> 31:02.000]  But in patient monitoring, what could I mean, you imagine the data that's collected, like, especially from an AI perspective, you know, obviously an algorithm is built by data.
[31:02.000 --> 31:07.000]  Your machines collect probably more data than a lot of med tech products.
[31:08.000 --> 31:17.000]  Can you imagine how that data could be used to improve workflow, patient experience, diagnosis?
[31:17.000 --> 31:22.000]  And, you know, so that there's I can imagine that's why I ask that question to you guys.
[31:22.000 --> 31:29.000]  In radiology, AI is probably one of the more advanced areas, but in patient monitoring, it's probably not as advanced.
[31:30.000 --> 31:37.000]  So I just wonder what would be that blue sky thinking of what I could do for your patient monitors?
[31:38.000 --> 31:41.000]  I would say I think I think, you know, you're right.
[31:41.000 --> 31:43.000]  Big data. Again, we come back to connectivity.
[31:43.000 --> 31:45.000]  We're talking big data.
[31:45.000 --> 31:58.000]  I think from machine learning and understanding all of the data that happens with patients, primarily post surgery, it's probably going to be looking at all of those results.
[31:58.000 --> 32:01.000]  And linking them to pre surgery.
[32:01.000 --> 32:04.000]  What was the patient like pre surgery?
[32:04.000 --> 32:05.000]  You know, what was the condition?
[32:05.000 --> 32:08.000]  What were they facing and actually seeing some trends there?
[32:08.000 --> 32:15.000]  Well, actually, you know, when we saw this before, this is how we looked after this patient and the outcomes were perfect.
[32:15.000 --> 32:19.000]  Whereas sometimes, you know, I'm not saying it's guesswork by far.
[32:19.000 --> 32:21.000]  These guys are very, very clever people.
[32:21.000 --> 32:33.000]  But actually, the data is going to secure a little bit of confidence in people in terms of actually how they look after patients through their, you know, perioperative sort of journey.
[32:33.000 --> 32:35.000]  I agree.
[32:35.000 --> 32:37.000]  And it's early intervention as well.
[32:37.000 --> 32:46.000]  So the more time that people are monitored, you know, before post, that's really going to aid with their care.
[32:46.000 --> 32:48.000]  Early intervention does save lives.
[32:48.000 --> 32:50.000]  And that's going to be helpful.
[32:50.000 --> 32:57.000]  But honestly, longer, longer, longer, longer term down there, you know, you just look at SkyFi and they've got so much right, you know.
[32:57.000 --> 33:02.000]  So we're all going to be in scanners, get stuff done, get out and it's all going to be done in day surgery.
[33:02.000 --> 33:06.000]  But that's a long way away.
[33:06.000 --> 33:13.000]  Actually, when you think when I think of what a mobile phone looked like 20 years ago and what it looks like today, obviously, it's become smaller.
[33:13.000 --> 33:16.000]  It's now you don't need a camera because you just have your phone.
[33:16.000 --> 33:21.000]  And, you know, it is very scary to think what's 10 years, 20 years down the line.
[33:21.000 --> 33:28.000]  But I think it's only going to help create efficiencies for the NHS with workforce shortages, etc.
[33:28.000 --> 33:30.000]  It's only going to be a good thing.
[33:30.000 --> 33:35.000]  So I'd like to thank you both for being on today's show, Dave and Chris.
[33:35.000 --> 33:41.000]  I think we've learned loads more about you guys, but also about the Patient Monitoring Special Focus Group.
[33:41.000 --> 33:44.000]  So a big thank you to Chris and David for joining us.
[33:44.000 --> 33:46.000]  And thank you to all of our listeners.
[33:46.000 --> 33:54.000]  Join Sally and I again next week when we will be speaking to Chris Taylor and Anu Hutasari about the Axrem Sustainability Special Focus Group.
Colab paid products - Cancel contracts here
Release notes 
close
more_horiz
Please follow our blog to see more information about new features, tips and tricks, and featured notebooks such as Analyzing a Bank Failure with Colab.

2025-05-12
Pro+ Subscriptions give 600 CCUs now: https://x.com/GoogleColab/status/1913312272967209357

Launched Colab's fastest deep learning accelerator: v6e-1 (Trillium) TPUs!! 2x the high bandwidth memory as v5e-1 (32GB) and a whopping peak rating of 918 BF16 TFLOPS (nearly 3x A100)!️ Try them out with JAX & Flax in our example notebook

Enabled UTF-8 mode by default for Python

Added CUDA.jl and Pluto.jl for Julia

Enabled OpenCL for LightGBM

Python package upgrades

accelerate 1.5.2 -> 1.6.0
bigframes 1.42.0 -> 2.3.0
bokeh 3.6.3 -> 3.7.2
cudf-cu12 24.12.0 -> 25.2.1
google-genai 1.9.0 -> 1.14.0
highspy 1.9.0 -> 1.10.0
holidays 0.69 -> 0.72
huggingface-hub 0.30.1 -> 0.31.1
lxml 5.3.1 -> 5.4.0
nx_cugraph_cu12 24.12.0 -> 25.2.1
openai 1.70.0 -> 1.78.0
peft 0.14.0 -> 0.15.2
polars 1.14.0 -> 1.21.0
pyomo 6.8.2 -> 6.9.2
scipy 1.14.1 -> 1.15.3
sentence-transformers 3.4.1 -> 4.1.0
transformers 4.50.3 -> 4.51.3
urllib3 2.3.0 -> 2.4.0
xarray 2025.1.2 -> 2025.3.1
Python package inclusions

dataproc-spark-connect 0.7.3
libpysal 4.13.0
lightgbm
stumpy 1.13.0
torchao 0.10.0
torchtune 0.6.1
tsfresh 0.21.0
2025-04-09
Colab Data Science Agent launched!
Julia language runtimes available with GPU and TPU support GitHub.
Now your popular Kaggle datasets are cached for quick retrieval.
Upgraded Colab runtimes to Python 3.11.
Python package upgrades

bigframes 1.29.0 -> 1.42.0
TensorFlow 2.17.1 -> 2.18.0
tensorboard 2.17.1 -> 2.18.0
keras 3.5.0 -> 3.8.0
torch 2.5.1 -> 2.6.0
torchaudio 2.5.1 -> 2.6.0
torchvision 0.20.1 -> 0.21.0
fastai 2.7.18 -> 2.7.19
ipykernel 5.5.6 -> 6.17.1
google-genai 0.3.0 -> 1.9.0
google-auth 2.27.0 -> 2.38.0
Tornado 6.3.3 -> 6.4.2
jax 0.4.33 -> 0.5.2
accelerate 1.2.1 -> 1.5.2
transformers 4.47.1 -> 4.50.3
openai 1.57.4 -> 1.70.0
kagglehub 0.3.6 -> 0.3.11
earthengine-api 1.4.3 -> 1.5.9
google-cloud-bigquery 3.29.0 -> 3.31.0
bigquery-magics 0.8.1 -> 0.9.0
NumPy 1.26.4 -> 2.0.2
Python package inclusions

cuml-cu12 24.12.0
2025-01-13
Released version 1.2.0 of the (Open in Colab Chrome Extension).
Released minimizable comments with indicators next to cell.
TPU v5e-1 Runtimes are now available for selection (tweet).
GPU prices were decreased (tweet).
Python package upgrades

accelerate 1.1.1 -> 1.2.1
aiohttp 3.10.10 -> 3.11.11
altair 4.2.2 -> 5.5.0
bigframes 1.25.0 -> 1.29.0
cmake 3.30.5 -> 3.31.2
cvxpy 1.5.3 -> 3.6.0
earthengine-api 1.2.0 -> 1.4.3
folium 0.18.0 -> 0.19.3
holidays 0.60 -> 0.63
huggingface-hub 0.26.2 -> 0.27.0
jsonpickle 3.4.2 -> 4.0.1
kagglehub 0.3.3 -> 0.3.6
keras 3.4.1 -> 3.5.0
matplotlib 3.8.0 -> 3.10.0
openai 1.54.3 -> 1.57.4
pymc 5.18.0 -> 5.19.1
safetensors 0.4.5 -> 0.5.0
scikit-image 0.24.0 -> 0.25.0
scikit-learn 1.5.2 -> 1.6.0
sentence-transformers 3.2.1 -> 3.3.1
tensorflow 2.17.0 -> 2.17.1
torch 2.5.0 -> 2.5.1
torchaudio 2.5.0 -> 2.5.1
torchvision 0.20.0 -> 0.20.1
transformers 4.46.2 -> 4.47.1
wandb 0.18.6 -> 0.19.1
xarray 2024.10.0 -> 2024.11.0
Python package inclusions

google-genai 0.3.0
2024-11-11
Users can now import Gemini API keys from AI Studio into their user secrets, all in Colab (tweet).
Increased limit to 1000 characters for requests to Gemini in Chat and Generate windows.
Improved saving notebook to GitHub flow.
Updated Gemini spark icon to be colorful.
uv is pre-installed on the PATH for faster package installs.
Fixed bugs
Dropdown text for GitHub repository not visible #4901.
Pre-installed California housing dataset README not correct #4862.
Backend execution error for scheduled notebook #4850.
Drive File Stream issues #3441.
Linking to the signup page does not preserve the authuser parameter.
Error messages in Gemini chat are not polished.
Clicking in Gemini chat feedback causes jitters the UI.
Hovering over a table of contents entry would show the menu icons for all entries.
Surveys display over open dialogs.
Playground mode banner not shown on mobile.
Python package upgrades

accelerate 0.34.2 -> 1.1.1
arviz 0.19.0 -> 0.20.0
bigframes 1.18.0 -> 1.25.0
bigquery-magics 0.2.0 -> 0.4.0
bokeh 3.4.3 -> 3.6.1
blosc 2.0.0 -> 2.7.1
cloudpickle 2.2.1 -> 3.1.0
cudf-cu12 24.4.1 -> 24.10.1
dask 2024.8.0 -> 24.10.0
debugpy 1.6.6 -> 1.8.0
earthengine-api 1.0.0 -> 1.2.0
folium 0.17.0 -> 0.18.0
gscfs 2024.6.1 -> 2024.10.0
geemap 0.34.3 -> 0.35.1
holidays 0.57 -> 0.60
huggingface-hub 0.24.7 -> 0.26.2
kagglehub 0.3.0 -> 0.3.3
lightgbm 4.4.0 -> 4.5.0
lxml 4.9.4 -> 5.3.0
matplotlib 3.7.1 -> 3.8.0
mizani 0.11.4 -> 0.13.0
networkx 3.3 -> 3.4.2
nltk 3.8.1 -> 3.9.1
pandas 2.1.4 -> 2.2.2
pillow 10.4.0 -> 11.0.0
plotnine 0.13.6 -> 0.14.1
polars 1.6.0 -> 1.9.0
protobuf 3.20.3 -> 4.25.5
pyarrow 14.0.2 -> 17.0.0
pydrive2 1.20.0 -> 1.21.1
pymc 5.16.2 -> 5.18.0
torch 2.4.1 -> 2.5.0
torchaudio 2.4.1 -> 2.5.0
torchvision 0.19.1 -> 0.20.0
transformers 4.44.2 -> 4.46.2
xarray 2024.9.0 -> 2024.10.0
Python package inclusions

diffusers 0.31.0
gitpython 3.1.43
langchain 0.3.7
openai 1.54.3
pygit2 1.16.0
pyspark 3.5.3
sentence-transformers 3.2.1
timm 1.0.11
wandb 0.18.6
Library and driver upgrades

drivefs upgraded from 89.0.2 to 98.0.0
2024-09-23
Improved code snippet search
Updated Marketplace image and public local runtime container
Improved the look-and-feel of interactive form dropdowns and checkboxes
Fixed bugs
activating the skip link caused the notebook to scroll out of view
toggling a checkbox too much caused the page to crash
lightning fast drags could cause orphaned tabs
custom widgets snippet would show for local runtimes
Python package upgrades

accelerate 0.32.1 -> 0.34.2
arviz 0.18.0 -> 0.19
autograd 1.6.2 -> 1.7.0
bigframes 1.14.0 -> 1.18.0
dask 2024.7.1 -> 2024.8.0
distributed 2024.7.1 -> 2024.8.0
duckdb 0.10.3 -> 1.1.0
earthengine-api 0.1.416 -> 1.0.0
flax 0.8.4 -> 0.8.5
gdown 5.1.0 -> 5.2.0
geemap 0.33.1 -> 0.34.3
geopandas 0.14.4 -> 1.0.1
google-cloud-aiplatform 1.59.0 -> 1.67.1
google-cloud-bigquery-storage 2.25.0 -> 2.26.0
holidays 0.54 -> 0.57
huggingface-hub 0.23.5 -> 0.24.7
ibis-framework 8.0.0 -> 9.2.0
jax 0.4.26 -> 0.4.33
jaxlib 0.4.26 -> 0.4.33
kagglehub 0.2.9 -> 0.3.0
lightgbm 4.4.0 -> 4.5.0
matplotlib-venn 0.11.10 -> 1.1.1
mizani 0.9.3 -> 0.11.4
Pillow 9.4.0 -> 10.4.0
plotly 5.15.0 -> 5.24.1
plotnine 0.12.4 -> 0.13.6
polars 0.20.2 -> 1.6.0
progressbar2 4.2.0 -> 4.5.0
PyDrive2 1.6.3 -> 1.20.0
pymc 5.10.4 -> 5.16.2
pytensor 2.18.6 -> 2.25.4
scikit-image 0.23.2 -> 0.24.0
scikit-learn 1.3.2 -> 1.5.2
torch 2.3.1 -> 2.4.1
torchaudio 2.3.1 -> 2.4.1
torchvision 0.18.1 -> 0.19.1
transformers 4.42.4 -> 4.44.2
urllib3 2.0.7 -> 2.2.3
xarray 2024.6.0 -> 2024.9.0
Python package inclusions

bigquery-magics 0.2.0
2024-08-20
TPU memory usage and utilization can now be checked with !tpu-info

Gemini Chat responses are now grounded in relevant sources

Added a new "Create Gemini API key" link in the user secrets panel

Added a new "Gemini: Creating a prompt" snippet and touched up the existing "Gemini: Connecting to Gemini" snippet

Added the ability to specify custom placeholder text for various interactive form params (see examples)

Keyboard navigation a11y improvements to comments UI

Various minor rendering improvements to interactive forms UI

A11y improvements for the run button and header

Updated tooltip styling

A11y improvements for the file browser’s disk usage bar

On mobile, tooltips now trigger on long press

On mobile, release notes updates will no longer display automatically

Python package upgrades

astropy 5.3.4 -> 6.1.2
bigframes 1.11.1 -> 1.14.0
bokeh 3.3.4 -> 3.4.3
dask 2023.8.1 -> 2024.7.1
earthengine-api 0.1.412 -> 0.1.416
geopandas 0.13.2 -> 0.14.4
kagglehub 0.2.8 -> 0.2.9
keras 2.15.0 -> 3.4.1
lightgbm 4.1.0 -> 4.4.0
malloy 2023.1067 -> 2024.1067
numba 0.58.1 -> 0.60.0
numpy 1.25.2 -> 1.26.4
opencv-python 4.8.0.76 -> 4.10.0.84
pandas 2.0.3 -> 2.1.4
pandas-gbq 0.19.2 -> 0.23.1
panel 1.3.8 -> 1.4.5
requests 2.31.0 -> 2.32.3
scikit-learn 1.2.2. -> 1.3.2
scipy 1.11.4 -> 1.13.1
tensorboard 2.15.2 -> 2.17.0
tensorflow 2.15.0 -> 2.17.0
tf-keras 2.15.1 -> 2.17.0
xarray 2023.7.0 -> 2024.6.0
xgboost 2.0.3 -> 2.1.1
Python package inclusions

einops 0.8.0
2024-07-22
You can now embed Google sheets directly into Colab to streamline interactions with data with InteractiveSheet.

Example:

from google.colab import sheets
sh = sheets.InteractiveSheet()
df = sh.as_df()
Fixed multiple rendering bugs in cell editors with wide text content (i.e. text is no longer hidden or clipped)

Fixed multiple accessibility issues in Colab's comments feature (e.g. proper keyboard focus management, added accessibility landmarks, etc)

Fixed bug where AI code generation would fail for extremely long broken code snippets

Fixed multiple scrollbar bugs in the user secrets panel

Added the ability for workspace admin to purchase Colab Pro and Pro+ Subscriptions for users

Fixed bug where user secrets couldn’t be moved to a tab

Fixed several focus management accessibility issues in tabs, the table of contents, the left toolbar, and the run button

Fixed bug where overflowing cells may be omitted when pasting from Google Sheets

Fixed bug where the generate code button did not activate on touch

Python package upgrades

bigframes 1.9.0 -> 1.11.1
cvxpy 1.3.4 -> 1.5.2
earthengine-api 0.1.408 -> 0.1.412
google-api-core 2.11.1 -> 2.19.1
google-api-python-client 2.84.0 -> 2.137.0
google-cloud-aiplatform 1.56.0 -> 1.59.0
google-cloud-bigquery 3.21.0 -> 3.25.0
google-cloud-core 2.3.3 -> 2.4.1
google-cloud-datastore 2.15.2 -> 2.19.0
google-cloud-firestore 2.11.1 -> 2.16.1
google-cloud-functions 1.13.3 -> 1.16.4
google-generativeai 0.5.4 -> 0.7.2
kagglehub 0.2.5 -> 0.2.8
pip 23.1.2 -> 24.1.2
setuptools 67.7.2 -> 71.0.4
sympy 1.12.1 -> 1.13.1
torch 2.3.0 -> 2.3.1
transformers 4.41.2 -> 4.42.4
Python package inclusions

accelerate 0.32.1
2024-06-18
Inline AI completions are now available to users on the free-of-charge tier

Reduced latency for LSP and terminal connections

Improved quality of inline completions

Visual improvements to switch controls across Colab

Various bug fixes, performance and a11y improvements to the user secrets panel

Improved tooltip UX behavior

Improved behavior when copying data from Google Sheets and pasting in Colab

Scroll to cell fixes for single tabbed view and jump to cell command

Improved tab header behavior

A11y improvements for notebook-focused cells

Python package upgrades

torch 2.2.1 -> 2.3.0
torchaudio 2.2.1 -> 2.3.0
torchvision 0.17.1 -> 0.18.0
torchtext 0.17.1 -> 0.18.0
google-cloud-aiplatform 1.51.0 -> 1.56.0
bigframes 1.5.0 -> 1.8.0
regex 2023.12.25 -> 2024.5.15
2024-05-13
Code actions are now supported to automatically improve and refactor code. Code actions can be triggered by the keyboard shortcut "Ctrl/⌘ + ."

Python package upgrades

bigframes 1.0.0 -> 1.5.0
google-cloud-aiplatform 1.47.0 -> 1.51.0
jax[tpu] 0.4.23 -> 0.4.26
Python package inclusions

cudf 24.4.1
2024-04-15
TPU v2 runtime is now available

L4 runtime is now available for paid users

New distributed fine-tuning Gemma tutorial on TPUs (GitHub)

Symbol rename is now supported with keyboard shortcut F2

Fixed bug causing inability to re-upload deleted files

Fixed breaking bug in colabtools %upload_files_async

Added syntax highlighting to %%writefile cells

Cuda dependencies that come with Torch are cached for faster downloads for packages that require Torch and its dependencies (GitHub issue)

Python package upgrades

bigframes 0.24.0 -> 1.0.0
duckdb 0.9.2 -> 0.10.1
google-cloud-aiplatform 1.43.0 -> 1.47.0
jax 0.4.23 -> 0.4.26
2024-03-13
Fixed bug that sometimes caused UserSecrets to move / disappear

Improved messaging for mounting drive in an unsupported environment (GitHub issue)

Python package upgrades

torch 2.1.0 -> 2.2.1
torchaudio 2.1.0 -> 2.2.1
torchvision 0.16.0 -> 0.17.1
torchtext 0.16.0 -> 0.17.1
PyMC 5.7.2 -> 5.10.4
BigFrames 0.21.0 -> 0.24.0
google-cloud-aiplatform 1.42.1 -> 1.43.0
tornado 6.3.2 -> 6.3.3
2024-02-21
Try out Gemma on Colab!

Allow unicode in form text inputs

Display documentation and link to source when displaying functions

Display image-like ndarrays as images

Improved UX around quick charts and execution error suggestions

Released Marketplace image for the month of February (GitHub issue)

Python package upgrades

bigframes 0.19.2 -> 0.21.0
regex 2023.6.3 -> 2023.12.25
spacy 3.6.1 -> 3.7.4
beautifulsoup4 4.11.2 -> 4.12.3
tensorflow-probability 0.22.0 -> 0.23.0
google-cloud-language 2.9.1 -> 2.13.1
google-cloud-aiplatform 1.39.0 -> 1.42.1
transformers 4.35.2 -> 4..37.2
pyarrow 10.0.1 -> 14.0.2
2024-01-29
New Kaggle Notebooks <> Colab updates! Now you can:

Import directly from Colab without having to download/re-upload
Upload via link, by pasting Google Drive or Colab URLs
Export & run Kaggle Notebooks on Colab with 1 click
Try these notebooks that talk to Gemini:

Gemini and Stable Diffusion
Learning with Gemini and ChatGPT
Talk to Gemini with Google's Speech to Text API
Sell lemonade with Gemini and Sheets
Generate images with Gemini and Vertex
Python package upgrades

google-cloud-aiplatform 1.38.1 -> 1.39.0
bigframes 0.18.0 -> 0.19.2
polars 0.17.3 -> 0.20.2
gdown 4.6.6 -> 4.7.3 (GitHub issue)
tensorflow-hub 0.15.0 -> 0.16.0
flax 0.7.5 -> 0.8.0
Python package inclusions

sentencepiece 0.1.99
2024-01-08
Avoid nested scrollbars for large outputs by using google.colab.output.no_vertical_scroll() Example notebook

Fix bug where downloading models from Hugging Face could freeze

Python package upgrades

huggingface-hub 0.19.4 -> 0.20.2
bigframes 0.17.0 -> 0.18.0
2023-12-18
Expanded access to AI coding has arrived in Colab across 175 locales for all tiers of Colab users

Improvements to display of ML-based inline completions (for eligible Pro/Pro+ users)

Started a series of notebooks highlighting Gemini API capabilities

Enable ⌘/Ctrl+L to select the full line in an editor

Fixed bug where we weren't correctly formatting output from multiple execution results

Python package upgrades

CUDA 11.8 to CUDA 12.2
tensorflow 2.14.0 -> 2.15.0
tensorboard 2.14.0 -> 2.15.0
keras 2.14.0 -> 2.15.0
Nvidia drivers 525.105.17 -> 535.104.05
tensorflow-gcs-config 2.14.0 -> 2.15.0
bigframes 0.13.0 -> 0.17.0
geemap 0.28.2 -> 0.29.6
pyarrow 9.0.0 -> 10.0.1
google-generativeai 0.2.2 -> 0.3.1
jax 0.4.20 —> 0.4.23
jaxlib 0.4.20 —> 0.4.23
Python package inclusions

kagglehub 0.1.4
google-cloud-aiplatform 1.38.1
2023-11-27
Removed warning when calling await to make it render as code

Added "Run selection" to the cell context menu

Added highlighting for the %%python cell magic

Launched AI coding features for Pro/Pro+ users in more locales

Python package upgrades

bigframes 0.12.0 -> 0.13.0
Python package inclusions

transformers 4.35.2
google-generativeai 0.2.2
2023-11-08
Launched Secrets, for safe storage of private keys on Colab (tweet)

Fixed issue where TensorBoard would not load (#3990)

Python package upgrades

lightgbm 4.0.0 -> 4.1.0
bigframes 0.10.0 -> 0.12.0
bokeh 3.2.2 -> 3.3.0
duckdb 0.8.1 -> 0.9.1
numba 0.56.4 -> 0.58.1
tweepy 4.13.0 -> 4.14.0
jax 0.4.16 -> 0.4.20
jaxlib 0.4.16 -> 0.4.20
2023-10-23
Updated the Open notebook dialog for better usability and support for smaller screen sizes

Added smart paste support for data from Google Sheets for R notebooks

Enabled showing release notes in a tab

Launched AI coding features for Pro/Pro+ users in Australia 🇦🇺 Canada 🇨🇦 India 🇮🇳 and Japan 🇯🇵 (tweet)

Python package upgrades

earthengine-api 0.1.357 -> 0.1.375
flax 0.7.2 -> 0.7.4
geemap 0.27.4 -> 0.28.2
jax 0.4.14 -> 0.4.16
jaxlib 0.4.14 -> 0.4.16
keras 2.13.1 -> 2.14.0
tensorboard 2.13.0 -> 2.14.1
tensorflow 2.13.0 -> 2.14.0
tensorflow-gcs-config 2.13.0 -> 2.14.0
tensorflow-hub 0.14.0 -> 0.15.0
tensorflow-probability 0.20.1 -> 0.22.0
torch 2.0.1 -> 2.1.0
torchaudio 2.0.2 -> 2.1.0
torchtext 0.15.2 -> 0.16.0
torchvision 0.15.2 -> 0.16.0
xgboost 1.7.6 -> 2.0.0
Python package inclusions

bigframes 0.10.0
malloy 2023.1056
2023-09-22
Added the ability to scope an AI generated suggestion to a specific Pandas dataframe (tweet)
Added Colab link previews to Docs (tweet)
Added smart paste support for data from Google Sheets
Increased font size of dropdowns in interactive forms
Improved rendering of the notebook when printing
Python package upgrades
tensorflow 2.12.0 -> 2.13.0
tensorboard 2.12.3 -> 2.13.0
keras 2.12.0 -> 2.13.1
tensorflow-gcs-config 2.12.0 -> 2.13.
scipy 1.10.1-> 1.11.2
cython 0.29.6 -> 3.0.2
Python package inclusions
geemap 0.26.0
2023-08-18
Added "Change runtime type" to the menu in the connection button
Improved auto-reconnection to an already running notebook (#3764)
Increased the specs of our highmem machines for Pro users
Fixed add-apt-repository command on Ubuntu 22.04 runtime (#3867)
Python package upgrades
bokeh 2.4.3 -> 3.2.2
cmake 3.25.2 -> 3.27.2
cryptography 3.4.8 -> 41.0.3
dask 2022.12.1 -> 2023.8.0
distributed 2022.12.1 -> 2023.8.0
earthengine-api 0.1.358 -> 0.1.364
flax 0.7.0 -> 0.7.2
ipython-sql 0.4.0 -> 0.5.0
jax 0.4.13 -> 0.4.14
jaxlib 0.4.13 -> 0.4.14
lightgbm 3.3.5 -> 4.0.0
mkl 2019.0 -> 2023.2.0
notebook 6.4.8 -> 6.5.5
numpy 1.22.4 -> 1.23.5
opencv-python 4.7.0.72 -> 4.8.0.76
pillow 8.4.0 -> 9.4.0
plotly 5.13.1 -> 5.15.0
prettytable 0.7.2 -> 3.8.0
pytensor 2.10.1 -> 2.14.2
spacy 3.5.4 -> 3.6.1
statsmodels 0.13.5 -> 0.14.0
xarray 2022.12.0 -> 2023.7.0
Python package inclusions
PyDrive2 1.6.3
2023-07-21
Launched auto-plotting for dataframes, available using the chart button that shows up alongside datatables (post)


Added a menu to the table of contents to support running a section or collapsing/expanding sections (post)


Added an option to automatically run the first cell or section, available under Edit -> Notebook settings (post)


Launched Pro/Pro+ to Algeria, Argentina, Chile, Ecuador, Egypt, Ghana, Kenya, Malaysia, Nepal, Nigeria, Peru, Rwanda, Saudi Arabia, South Africa, Sri Lanka, Tunisia, and Ukraine (tweet)

Added a command, "Toggle tab moves focus" for toggling tab trapping in the editor (Tools -> Command palette, "Toggle tab moves focus")

Fixed issue where files.upload() was sometimes returning an incorrect filename (#1550)

Fixed f-string syntax highlighting bug (#3802)

Disabled ambiguous characters highlighting for commonly used LaTeX characters (#3648)

Upgraded Ubuntu from 20.04 LTS to 22.04 LTS

Updated the Colab Marketplace VM image

Python package upgrades:

autograd 1.6.1 -> 1.6.2
drivefs 76.0 -> 77.0
flax 0.6.11 -> 0.7.0
earthengine-api 0.1.357 -> 0.1.358
GDAL 3.3.2->3.4.3
google-cloud-bigquery-storage 2.20.0 -> 2.22.2
gspread-dataframe 3.0.8 -> 3.3.1
holidays 0.27.1 -> 0.29
jax 0.4.10 -> jax 0.4.13
jaxlib 0.4.10 -> jax 0.4.13
jupyterlab-widgets 3.0.7 -> 3.0.8
nbformat 5.9.0 -> 5.9.1
opencv-python-headless 4.7.0.72 -> 4.8.0.74
pygame 2.4.0 -> 2.5.0
spacy 3.5.3 -> 3.5.4
SQLAlchemy 2.0.16 -> 2.0.19
tabulate 0.8.10 -> 0.9.0
tensorflow-hub 0.13.0 -> 0.14.0
2023-06-23
Launched AI coding features to subscribed users starting with Pro+ users in the US (tweet, post)
Added the Kernel Selector in the Notebook Settings (tweet)
Fixed double space trimming issue in markdown #3766
Fixed run button indicator not always centered #3609
Fixed inconsistencies for automatic indentation on multi-line #3697
Upgraded Python from 3.10.11 to 3.10.12
Python package updates:
duckdb 0.7.1 -> 0.8.1
earthengine-api 0.1.350 -> 0.1.357
flax 0.6.9 -> 0.6.11
google-cloud-bigquery 3.9.0 -> 3.10.0
google-cloud-bigquery-storage 2.19.1 -> 2.20.0
grpcio 1.54.0 -> 1.56.0
holidays 0.25 -> 0.27.1
nbformat 5.8.0 -> 5.9.0
prophet 1.1.3 -> 1.1.4
pydata-google-auth 1.7.0 -> 1.8.0
spacy 3.5.2 -> 3.5.3
tensorboard 2.12.2 -> 2.12.3
xgboost 1.7.5 -> 1.7.6
Python package inclusions:
gcsfs 2023.6.0
geopandas 0.13.2
google-cloud-bigquery-connection 1.12.0
google-cloud-functions 1.13.0
grpc-google-iam-v1 0.12.6
multidict 6.0.4
tensorboard-data-server 0.7.1
2023-06-02
Released the new site colab.google
Published Colab's Docker runtime image to us-docker.pkg.dev/colab-images/public/runtime (tweet, instructions)
Launched support for Google children accounts (tweet)
Launched DagsHub integration (tweet, post)
Upgraded to Monaco Editor Version 0.37.1
Fixed various Vim keybinding bugs
Fixed issue where the N and P letters sometimes couldn't be typed (#3664)
Fixed rendering support for compositional inputs (#3660, #3679)
Fixed lag in notebooks with lots of cells (#3676)
Improved support for R by adding a Runtime type notebook setting (Edit -> Notebook settings)
Improved documentation for connecting to a local runtime (Connect -> Connect to a local runtime)
Python package updates:
holidays 0.23 -> 0.25
jax 0.4.8 -> 0.4.10
jaxlib 0.4.8 -> 0.4.10
pip 23.0.1 -> 23.1.2
tensorflow-probability 0.19.0 -> 0.20.1
torch 2.0.0 -> 2.0.1
torchaudio 2.0.1 -> 2.0.2
torchdata 0.6.0 -> 0.6.1
torchtext 0.15.1 -> 0.15.2
torchvision 0.15.1 -> 0.15.2
tornado 6.2 -> 6.3.1
2023-05-05
Released GPU type selection for paid users, allowing them to choose a preferred NVidia GPU
Upgraded R from 4.2.3 to 4.3.0
Upgraded Python from 3.9.16 to 3.10.11
Python package updates:
attrs 22.2.0 -> attrs 23.1.0
earthengine-api 0.1.349 -> earthengine-api 0.1.350
flax 0.6.8 -> 0.6.9
grpcio 1.53.0 -> 1.54.0
nbclient 0.7.3 -> 0.7.4
tensorflow-datasets 4.8.3 -> 4.9.2
termcolor 2.2.0 -> 2.3.0
zict 2.2.0 -> 3.0.0
2023-04-14
Python package updates:
google-api-python-client 2.70.0 -> 2.84.0
google-auth-oauthlib 0.4.6 -> 1.0.0
google-cloud-bigquery 3.4.2 -> 3.9.0
google-cloud-datastore 2.11.1 -> 2.15.1
google-cloud-firestore 2.7.3 -> 2.11.0
google-cloud-language 2.6.1 -> 2.9.1
google-cloud-storage 2.7.0 -> 2.8.0
google-cloud-translate 3.8.4 -> 3.11.1
networkx 3.0 -> 3.1
notebook 6.3.0 -> 6.4.8
jax 0.4.7 -> 0.4.8
pandas 1.4.4 -> 1.5.3
spacy 3.5.1 -> 3.5.2
SQLAlchemy 1.4.47 -> 2.0.9
xgboost 1.7.4 -> 1.7.5
2023-03-31
Improve bash ! syntax highlighting (GitHub issue)
Fix bug where VIM keybindings weren't working in the file editor
Upgraded R from 4.2.2 to 4.2.3
Python package updates:
arviz 0.12.1 --> 0.15.1
astropy 4.3.1 --> 5.2.2
dopamine-rl 1.0.5 --> 4.0.6
gensim 3.6.0 --> 4.3.1
ipykernel 5.3.4 -> 5.5.6
ipython 7.9.0 -> 7.34.0
jax 0.4.4 -> 0.4.7
jaxlib 0.4.4 -> 0.4.7
jupyter_core 5.2.0 -> 5.3.0
keras 2.11.0 -> 2.12.0
lightgbm 2.2.3 -> 3.3.5
matplotlib 3.5.3 -> 3.7.1
nltk 3.7 -> 3.8.1
opencv-python 4.6.0.66 -> 4.7.0.72
plotly 5.5.0 -> 5.13.1
pymc 4.1.4 -> 5.1.2
seaborn 0.11.2 -> 0.12.2
spacy 3.4.4 -> 3.5.1
sympy 1.7.1 -> 1.11.1
tensorboard 2.11.2 -> 2.12.0
tensorflow 2.11.0 -> 2.12.0
tensorflow-estimator 2.11.0 -> 2.12.0
tensorflow-hub 0.12.0 -> 0.13.0
torch 1.13.1 -> 2.0.0
torchaudio 0.13.1 -> 2.0.1
torchtext 0.14.1 -> 0.15.1
torchvision 0.14.1 -> 0.15.1
2023-03-10
Added the Colab editor shortcuts example notebook
Fixed triggering of @-mention and email autocomplete for large comments (GitHub issue)
Added View Resources to the Runtime menu
Made file viewer images fit the view by default, resizing to original size on click
When in VIM mode, enable copy as well as allowing propagation to monaco-vim to escape visual mode (GitHub issue)
Upgraded CUDA 11.6.2 -> 11.8.0 and cuDNN 8.4.0.27 -> 8.7.0.84
Upgraded Nvidia drivers 525.78.01 -> 530.30.02
Upgraded Python 3.8.10 -> 3.9.16
Python package updates:
beautifulsoup4 4.6.3 -> 4.9.3
bokeh 2.3.3 -> 2.4.3
debugpy 1.0.0 -> 1.6.6
Flask 1.1.4 -> 2.2.3
jax 0.3.25 -> 0.4.4
jaxlib 0.3.25 -> 0.4.4
Jinja2 2.11.3 -> 3.1.2
matplotlib 3.2.2 -> 3.5.3
nbconvert 5.6.1 -> 6.5.4
pandas 1.3.5 -> 1.4.4
pandas-datareader 0.9.0 -> 0.10.0
pandas-profiling 1.4.1 -> 3.2.0
Pillow 7.1.2 -> 8.4.0
plotnine 0.8.0 -> 0.10.1
scikit-image 0.18.3 -> 0.19.3
scikit-learn 1.0.2 -> 1.2.2
scipy 1.7.3 -> 1.10.1
setuptools 57.4.0 -> 63.4.3
sklearn-pandas 1.8.0 -> 2.2.0
statsmodels 0.12.2 -> 0.13.5
urllib3 1.24.3 -> 1.26.14
Werkzeug 1.0.1 -> 2.2.3
wrapt 1.14.1 -> 1.15.0
xgboost 0.90 -> 1.7.4
xlrd 1.2.0 -> 2.0.1
2023-02-17
Show graphs of RAM and disk usage in notebook toolbar
Copy cell links directly to the clipboard instead of showing a dialog when clicking on the link icon in the cell toolbar
Updated the Colab Marketplace VM image
Upgraded CUDA to 11.6.2 and cuDNN to 8.4.0.27
Python package updates:
tensorflow 2.9.2 -> 2.11.0
tensorboard 2.9.1 -> 2.11.2
keras 2.9.0 -> 2.11.0
tensorflow-estimator 2.9.0 -> 2.11.0
tensorflow-probability 0.17.0 -> 0.19.0
tensorflow-gcs-config 2.9.0 -> 2.11.0
earthengine-api 0.1.339 -> 0.1.341
flatbuffers 1.12 -> 23.1.21
platformdirs 2.6.2 -> 3.0.0
pydata-google-auth 1.6.0 -> 1.7.0
python-utils 3.4.5 -> 3.5.2
tenacity 8.1.0 -> 8.2.1
tifffile 2023.1.23.1 -> 2023.2.3
notebook 5.7.16 -> 6.3.0
tornado 6.0.4 -> 6.2
aiohttp 3.8.3 -> 3.8.4
charset-normalizer 2.1.1 -> 3.0.1
fastai 2.7.0 -> 2.7.1
soundfile 0.11.0 -> 0.12.1
typing-extensions 4.4.0 -> 4.5.0
widgetsnbextension 3.6.1 -> 3.6.2
pydantic 1.10.4 -> 1.10.5
zipp 3.12.0 -> 3.13.0
numpy 1.21.6 -> 1.22.4
drivefs 66.0 -> 69.0
gdal 3.0.4 -> 3.3.2 GitHub issue
Added libudunits2-dev for smoother R package installs GitHub issue
2023-02-03
Improved tooltips for pandas series to show common statistics about the series object
Made the forms dropdown behave like an autocomplete box when it allows input
Updated the nvidia driver from 460.32.03 to 510.47.03
Python package updates:
absl-py 1.3.0 -> 1.4.0
bleach 5.0.1 -> 6.0.0
cachetools 5.2.1 -> 5.3.0
cmdstanpy 1.0.8 -> 1.1.0
dnspython 2.2.1 -> 2.3.0
fsspec 2022.11.0 -> 2023.1.0
google-cloud-bigquery-storage 2.17.0 -> 2.18.1
holidays 0.18 -> 0.19
jupyter-core 5.1.3 -> 5.2.0
packaging 21.3 -> 23.0
prometheus-client 0.15.0 -> 0.16.0
pyct 0.4.8 -> 0.5.0
pydata-google-auth 1.5.0 -> 1.6.0
python-slugify 7.0.0 -> 8.0.0
sqlalchemy 1.4.46 -> 2.0.0
tensorflow-io-gcs-filesystem 0.29.0 -> 0.30.0
tifffile 2022.10.10 -> 2023.1.23.1
zipp 3.11.0 -> 3.12.0
Pinned sqlalchemy to version 1.4.46
2023-01-12
Added support for @-mention and email autocomplete in comments
Improved errors when GitHub notebooks can't be loaded
Increased color contrast for colors used for syntax highlighting in the code editor
Added terminal access for custom GCE VM runtimes
Upgraded Ubuntu from 18.04 LTS to 20.04 LTS (GitHub issue)
Python package updates:
GDAL 2.2.2 -> 2.2.3.
NumPy from 1.21.5 to 1.21.6.
attrs 22.1.0 -> 22.2.0
chardet 3.0.4 -> 4.0.0
cloudpickle 1.6.0 -> 2.2.0
filelock 3.8.2 -> 3.9.0
google-api-core 2.8.2 -> 2.11.0
google-api-python-client 1.12.11 -> 2.70.0
google-auth-httplib2 0.0.3 -> 0.1.0
google-cloud-bigquery 3.3.5 -> 3.4.1
google-cloud-datastore 2.9.0 -> 2.11.0
google-cloud-firestore 2.7.2 -> 2.7.3
google-cloud-storage 2.5.0 -> 2.7.0
holidays 0.17.2 -> holidays 0.18
importlib-metadata 5.2.0 -> 6.0.0
networkx 2.8.8 -> 3.0
opencv-python-headless 4.6.0.66 -> 4.7.0.68
pip 21.1.3 -> 22.04
pip-tools 6.2.0 -> 6.6.2
prettytable 3.5.0 -> 3.6.0
requests 2.23.0 -> 2.25.1
termcolor 2.1.1 -> 2.2.0
torch 1.13.0 -> 1.13.1
torchaudio 0.13.0 -> 0.13.1
torchtext 0.14.0-> 0.14.1
torchvision 0.14.0 -> 0.14.1
2022-12-06
Made fallback runtime version available until mid-December (GitHub issue)
Upgraded to Python 3.8 (GitHub issue)
Python package updates:
jax from 0.3.23 to 0.3.25, jaxlib from 0.3.22 to 0.3.25
pyarrow from 6.0.1 to 9.0.0
torch from 1.12.1 to 1.13.0
torchaudio from 0.12.1 to 0.13.0
torchvision from 0.13.1 to 0.14.0
torchtext from 0.13.1 to 0.14.0
xlrd from 1.1.0 to 1.2.0
DriveFS from 62.0.1 to 66.0.3
Made styling of markdown tables in outputs match markdown tables in text cells
Improved formatting for empty interactive table rows
Fixed syntax highlighting for variables with names that contain Python keywords (GitHub issue)
2022-11-11
Added more dark editor themes for Monaco (when in dark mode, "Editor colorization" appears as an option in the Editor tab of the Tools → Settings dialog)
Fixed bug where collapsed forms were deleted on mobile GitHub issue
Python package updates:
rpy2 from 3.4.0 to 3.5.5 (GitHub issue)
notebook from 5.5.0 to 5.7.16
tornado from 5.1.1 to 6.0.4
tensorflow_probability from 0.16.0 to 0.17.0
pandas-gbq from 0.13.3 to 0.17.9
protobuf from 3.17.3 to 3.19.6
google-api-core[grpc] from 1.31.5 to 2.8.2
google-cloud-bigquery from 1.21.0 to 3.3.5
google-cloud-core from 1.0.1 to 2.3.2
google-cloud-datastore from 1.8.0 to 2.9.0
google-cloud-firestore from 1.7.0 to 2.7.2
google-cloud-language from 1.2.0 to 2.6.1
google-cloud-storage from 1.18.0 to 2.5.0
google-cloud-translate from 1.5.0 to 3.8.4
2022-10-21
Launched a single-click way to get from BigQuery to Colab to further explore query results (announcement)
Launched Pro, Pro+, and Pay As You Go to 19 additional countries: Austria, Belgium, Bulgaria, Croatia, Cyprus, Czechia, Denmark, Estonia, Finland, Greece, Hungary, Latvia, Lithuania, Norway, Portugal, Romania, Slovakia, Slovenia, and Sweden (tweet)
Updated jax from 0.3.17 to 0.3.23, jaxlib from 0.3.15 to 0.3.22, TensorFlow from 2.8.2 to 2.9.2, CUDA from 11.1 to 11.2, and cuDNN from 8.0 to 8.1 (backend-info)
Added a readonly option to drive.mount
Fixed bug where Xarray was not working (GitHub issue)
Modified Markdown parsing to ignore block quote symbol within MathJax (GitHub issue)
2022-09-30
Launched Pay As You Go, allowing premium GPU access without requiring a subscription
Added vim and tcllib to our runtime image
Fixed bug where open files were closed on kernel disconnect (GitHub issue)
Fixed bug where the play button/execution indicator was not clickable when scrolled into the cell output (GitHub issue)
Updated the styling for form titles so that they avoid obscuring the code editor
Created a GitHub repo, backend-info, with the latest apt-list.txt and pip-freeze.txt files for the Colab runtime (GitHub issue)
Added files.upload_file(filename) to upload a file from the browser to the runtime with a specified filename
2022-09-16
Upgraded pymc from 3.11.0 to 4.1.4, jax from 0.3.14 to 0.3.17, jaxlib from 0.3.14 to 0.3.15, fsspec from 2022.8.1 to 2022.8.2
Modified our save flow to avoid persisting Drive filenames as titles in notebook JSON
Updated our Terms of Service
Modified the Jump to Cell command to locate the cursor at the end of the command palette input (Jump to cell in Tools → Command palette in a notebook with section headings)
Updated the styling of the Drive notebook comment UI
Added support for terminating your runtime from code: python from google.colab import runtime runtime.unassign()
Added regex filter support to the Recent notebooks dialog
Inline google.colab.files.upload JS to fix files.upload() not working (GitHub issue)
2022-08-26
Upgraded PyYAML from 3.13 to 6.0 (GitHub issue), drivefs from 61.0.3 to 62.0.1
Upgraded TensorFlow from 2.8.2 to 2.9.1 and ipywidgets from 7.7.1 to 8.0.1 but rolled both back due to a number of user reports (GitHub issue, GitHub issue)
Stop persisting inferred titles in notebook JSON (GitHub issue)
Fix bug in background execution which affected some Pro+ users (GitHub issue)
Fix bug where Download as .py incorrectly handled text cells ending in a double quote
Fix bug for Pro and Pro+ users where we weren't honoring the preference (Tools → Settings) to use a temporary scratch notebook as the default landing page
Provide undo/redo for scratch cells
When writing ipynb files, serialize empty multiline strings as [] for better consistency with JupyterLab
2022-08-11
Upgraded ipython from 5.5.0 to 7.9.0, fbprophet 0.7 to prophet 1.1, tensorflow-datasets from 4.0.1 to 4.6.0, drivefs from 60.0.2 to 61.0.3, pytorch from 1.12.0 to 1.12.1, numba from 0.51 to 0.56, and lxml from 4.2.0 to 4.9.1
Loosened our requests version requirement (GitHub issue)
Removed support for TensorFlow 1
Added Help → Report Drive abuse for Drive notebooks
Fixed indentation for Python lines ending in [
Modified styling of tables in Markdown to left-align them rather than centering them
Fixed special character replacement when copying interactive tables as Markdown
Fixed ansi 8-bit color parsing (GitHub issue)
Configured logging to preempt transitive imports and other loading from implicitly configuring the root logger
Modified forms to use a value of None instead of causing a parse error when clearing raw and numeric-typed form fields
2022-07-22
Update scipy from 1.4.1 to 1.7.3, drivefs from 59.0.3 to 60.0.2, pytorch from 1.11 to 1.12, jax & jaxlib from 0.3.8 to 0.3.14, opencv-python from 4.1.2.30 to 4.6.0.66, spaCy from 3.3.1 to 3.4.0, and dlib from 19.18.0 to 19.24.0
Fix Open in tab doc link which was rendering incorrectly (GitHub issue)
Add a preference for the default tab orientation to the Site section of the settings menu under Tools → Settings
Show a warning for USE_AUTH_EPHEM usage when running authenticate_user on a TPU runtime (code)
2022-07-01
Add a preference for code font to the settings menu under Tools → Settings
Update drivefs from 58.0.3 to 59.0.3 and spacy from 2.2.4 to 3.3.1
Allow display_data and execute_result text outputs to wrap, matching behavior of JupyterLab (does not affect stream outputs/print statements).
Improve LSP handling of some magics, esp. %%writefile (GitHub issue).
Add a FAQ entry about the mount Drive button behavior and include link buttons for each FAQ entry.
Fix bug where the notebook was sometimes hidden behind other tabs on load when in single pane view.
Fix issue with inconsistent scrolling when an editor is in multi-select mode.
Fix bug where clicking on a link in a form would navigate away from the notebook
Show a confirmation dialog before performing Replace all from the Find and replace pane.
2022-06-10
Update drivefs from 57.0.5 to 58.0.3 and tensorflow from 2.8.0 to 2.8.2
Support more than 100 repos in the GitHub repo selector shown in the open dialog and the clone to GitHub dialog
Show full notebook names on hover in the open dialog
Improve the color contrast for links, buttons, and the ipywidgets.Accordion widget in dark mode
2022-05-20
Support URL params for linking to some common pref settings: force_theme=dark, force_corgi_mode=1, force_font_size=14. Params forced by URL are not persisted unless saved using Tools → Settings.
Add a class markdown-google-sans to allow Markdown to render in Google Sans
Update monaco-vim from 0.1.19 to 0.3.4
Update drivefs from 55.0.3 to 57.0.5, jax from 0.3.4 to 0.3.8, and jaxlib from 0.3.2 to 0.3.7
2022-04-29
Added 🦀 mode (under Miscellaneous in Tools → Settings)
Added "Disconnect and delete runtime" option to the menu next to the Connect button
Improved rendering of filter options in an interactive table
Added git-lfs to the base image
Updated torch from 1.10.0 to 1.11.0, jupyter-core from 4.9.2 to 4.10.0, and cmake from 3.12.0 to 3.22.3
Added more details to our FAQ about unsupported uses (using proxies, downloading torrents, etc.)
Fixed issue with apt-get dependencies
2022-04-15
Add an option in the file browser to show hidden files.
Upgrade gdown from 4.2.0 to 4.4.0, google-api-core[grpc] from 1.26.0 to 1.31.5, and pytz from 2018.4 to 2022.1
2022-03-25
Launched Pro/Pro+ to 12 additional countries: Australia, Bangladesh, Colombia, Hong Kong, Indonesia, Mexico, New Zealand, Pakistan, Philippines, Singapore, Taiwan, and Vietnam
Added google.colab.auth.authenticate_service_account() to support using Service Account keys
Update jax from 0.3.1 to 0.3.4 & jaxlib from 0.3.0 to 0.3.2
Fixed an issue with Twitter previews of notebooks shared as GitHub Gists
2022-03-10
Launched Pro/Pro+ to 10 new countries: Ireland, Israel, Italy, Morocco, the Netherlands, Poland, Spain, Switzerland, Turkey, and the United Arab Emirates
Launched support for scheduling notebooks for Pro+ users
Fixed bug in interactive datatables where filtering by number did not work
Finished removing the python2 kernelspec
2022-02-25
Made various accessibility improvements to the header
Fix bug with forms run:auto where a form field change would trigger multiple runs
Minor updates to the bigquery example notebook and snippet
Include background execution setting in the sessions dialog for Pro+ users
Update tensorflow-probability from 0.15 to 0.16
Update jax from 0.2.25 to 0.3.1 & jaxlib from 0.1.71 to 0.3.0
2022-02-11
Improve keyboard navigation for the open dialog
Fix issue where nvidia-smi stopped reporting resource utilization for some users who were modifying the version of nvidia used
Update tensorflow from 2.7 to 2.8, keras from 2.7 to 2.8, numpy from 1.19.5 to 1.21.5, tables from 3.4.4 to 3.7.0
2022-02-04
Improve UX for opening content alongside your notebook, such as files opened from the file browser. This includes a multi-pane view and drag-drop support
Better Twitter previews when sharing example Colab notebooks and notebooks opened from GitHub Gists
Update pandas from 1.1.5 to 1.3.5
Update openpyxl from 2.5.9 to 3.0.0 and pyarrow from 3.0.0 to 6.0.0
Link to the release notes from the Help menu
2022-01-28
Add a copy button to data tables
Python LSP support for better completions and code diagnostics. This can be configured in the Editor Settings (Tools → Settings)
Update gspread examples in our documentation
Update gdown from 3.6 to 4.2
2022-01-21
New documentation for the google.colab package
Show GPU RAM in the resource usage tab
Improved security for mounting Google Drive which disallows mounting Drive from accounts other than the one currently executing the notebook
2022-01-14
Add a preference (Tools → Settings) to use a temporary scratch notebook as the default landing page
Fix bug where / and : weren't working in VIM mode
Update gspread from 3.0 to 3.4
Update the Colab Marketplace VM image
data_object
Variables
terminal
Terminal