PM5 obsolete?

[SWE192]

Hi,

I bought my Concept 2 RowERG D two years ago. It is, without a doubt, the one and only rower to buy. No regrets whatsoever!

However, even two years ago, the PM5 felt somewhat dusted. The display, the UI, and connectivity to common devices are archaic. It reminds me of my HP RPN calculator from the 1980ies. Now, every time I row I think the PM5 is more obsolete than many more modern devices that have already passed into digital Nirvana.

I wish so much for a better, larger, more appealing display (e-ink?), more intuitive UE-optimized menus, integration of a broader range of biometrics (e.g., SO2, ECG - just have a look at the amazing stuff MoniCardi in Finland is doing!), and better connectivity to smartwatches (Apple Watch!!) - all features already realized in a range of other devices. Those are my amateur's ideas - experts likely see a lot more opportunities.

Any thoughts on this?

[drluvguru]

Totally agree! The PM5 does feel outdated, especially compared to modern smartwatches and fitness trackers. A larger, more modern display and improved connectivity would make a huge difference. I'd love to see Concept 2 incorporate more advanced biometrics and seamless integration with popular devices like the Apple Watch. Maybe they'll take our feedback to heart and release a major update or new model?

[jamesg]

It has just two bearings and a flywheel, with a piece of industrial 1/4" chain to spin it. Almost a simple as a boat, with the advantage that we can't fall in. What less can we want?

[JaapvanE]

Full disclosure: I'm the lead developer of OpenRowingMonitor (ORM), an open source alternative for the PM5 (primarily aimed at machines that lack decent metrics). We had the opportunity to revisit any design decission C2 seems to have made (with the help of some great resources that paved the way). We had the benefit of looking at any monitor in the market (we are close friends with the EXR development team, which owns many machines). As we run on a full-fledged Raspberry Pi, we have a 4 core CPU, plenty of memory and access to incredible hardware.

So, see my comments as someone who followed in C2's footsteps, but was able to take a radically different approach at every turn (given our team, these are true debates and we are inclined to pushing the envelope).

However, even two years ago, the PM5 felt somewhat dusted. The display, the UI, and connectivity to common devices are archaic. It reminds me of my HP RPN calculator from the 1980ies. Now, every time I row I think the PM5 is more obsolete than many more modern devices that have already passed into digital Nirvana.

A key thing to understand here is the intended key usage scenario of the monitor. A PM5 isn't designed to be used in a home scenario. It is designed to be part of a fleet of machines in a boathouse or in a gym. Fleet owners absolutely despise maintenance of any kind, especially if it puts machines out of commission. So the PM5 is clunky but extremely robust. I've seen PM3's survive decades of abuse and neglect, which is impressive by its own right.

The PM5 is so energy efficient it doesn't need a power cord, making the RowErg an extremely flexible device as it requires no outlets. For ORM you need a serious batterypack or a powercord.

Moving to a touchscreen might make the design more beautiful, but it is much less robust. Our standard config is a Raspberry Pi 4 with 7 inch touchscreen in an ABS case. One bad hit on the case and the screen is a goner.

Please also realise that the commercial price of a PM5 is about €180. For a RPi4, ABS case and 7' touchscreen, I pay around €200. We support a Raspberry Pi Zero 2W (resulting in a price drop of about €80), but the price of the screen and case are still considerable.

The latest RP3 Aqua has a gorgeous tablet mounted on it, and I am curious how fragile it is. As the machine is much more expensive than a C2, you can guess where the cost originate.

I wish so much for a better, larger, more appealing display (e-ink?), more intuitive UE-optimized menus,

Again, this comes down to the usage scenarios. I think the key role of the PM5 is to provide essential information to the rower when jumping on it casually, but more advanced stuff should be done in the ErgData app. In that app you have the opportunity for more advanced scenarios.

Looking at ORM, we are moving the exact same way. As our UI is webbased, it is a lot more flexible, but we still move to the app. The key issue is that an app/phone is linked to an individual person, where a monitor is linked to a machine. When you want flexibility in a fleet scenario, you don't want to store any user data on the PM5. You want it on a USBDrive or app (guess what C2 does?).

integration of a broader range of biometrics (e.g., SO2, ECG - just have a look at the amazing stuff MoniCardi in Finland is doing!),

A key distinction here is whether you see the PM5 (or ORM) as a 'simple' data generator or as a central point of your data processing. Again, as the PM5 is tied to the machine, not to the user, it shouldn't contain personal data.

ErgData just records the basics. In the ORM team we had a lot of discussion about this, as we see a lot of room for improvement in terms of metrics structurally recorded. We record a lot more data from the machine per stroke (force curves, drive time, drive length, etc). However, the PM5 reports this data via bluetooth. With apps, you can record it all and get a lot of extra data. You just have to upload it to a party like RowsAndAll.com to be able to use it. Again, this fits in a world view that the PM5 is just there to determine key machine metrics and share it with more intelligent consumers.

A key outcome for ORM is that we are not the place to record additional data (like breathing rythm, etc..), as we can't extract that data from any of the devices reporting to us, even if we wanted to. For example, a Garmin Epix records this during a session, but it won't share it with data consumers. Nor do any other sensors like HRM straps report this. Even if we could get the data, there is no output format that can really capture it all: RowsAndAll's csv's capture truly anything rowing related, but miss a lot of health data like breathing rythm. Garmin fit files are the industry standard that can record all health data, but really lack the rich rowing data.

We've thus chosen to follow C2's path: we report our data nicely to Garmin and Polar via BLE and ANT+, and they incorporate it into their recorded session. We have more info we record and share with RowsAndAll, and hopefully someday people realize that merging data from multiple sources is interesting.

better connectivity to smartwatches (Apple Watch!!) - all features already realized in a range of other devices.

That is because Apple thinks they are rulers of the earth or something, and completely disregard any industry standard, thinking that everybody will adopt to their standard if they set it. At ORM we also looked at Apple and decided that we won't play along their arrogance. We, like C2, focus on industry standards.

In developping ORM we looked at the industry golden standards that Apple has chosen to completely disregard: BLE FTMS and ANT+. ANT+ is supported by Garmin, Polar and several other sportwatches out there. Use it daily, works brilliantly. FTMS is supported by EXR and KinoMap, and works great as well (use that daily on ORM to connect to EXR).

A PM5 supports ANT+ in both directions, so not much more you can ask for. On the bluetooth side, the PM5 doesn't support BLE FTMS. But their own protocol predates FTMS, is a thousand times better (as FTMS only reports time, distance, pace and power) and its specs are public. And C2 is market leader. ORM even emulates a PM5 as many rowing apps only understand it instead of FTMS.

Those are my amateur's ideas - experts likely see a lot more opportunities.

In all honesty, my only issue with the PM5/ErgData combo are three things

• I would like two or three simultanuous BLE interfaces running. I like to use ErgData to get a verified session or another app to extract all data, and EXR on my tablet to keep me occupied.
• ErgData should record much more data into its logs. It gets it via BLE, it even shows it in the GUI, why not record it? Now I need third-party apps and third party sites to get data presented that a PM5/ErgData already has.
• A better integration with key providers of training plans (training peaks, intervals.icu) and key consumers of data (RowsAndAll, training peaks, intervals.icu). RowErgs are not on an island, they are typically part of a coherent training plan stored on some website outside C2. Clising that loop would make life a lot easier. Garmin carries this heavy load now, and I'm still not certain it should keep on carrying it.

[Carl Watts]

I suggested that C2 make a "Smart Tach" probably 5 years ago now.

Use a self powered Bluetooth tach with battery backup and just run the whole "Monitor" on your smartphone as an App, but Concept 2 have not been interested.

I think its coming, the PM6 will be your smartphone or tablet, people want the connectivity and a colour screen, especially if C2 take the real time loop and virtual rowing to the next level.

C2 somehow need to offset the loss of money selling hardware and pull it in with software and a subscription service with some decent graphics. You get the basic monitor for free but you start paying for the add on's like the real time loop, a more pro version of the monitor that gives you performance analysis etc..

[stevegaspars]

people want the connectivity and a colour screen

I would like two or three simultanuous BLE interfaces running.

They include a smart phone holder as standard. It's not hard to connect it to the machine. The vast majority of people have one. Surely the device can take care of the extra BLE connections by rebroadcasting; the fancy graphics and metrics; the touch screen.

The main monitor doesn't need to satisfy these requirements. Keep it simple, low power, easily repairable and swappable.

[Sakly]

I absolutely like the style how C2 provides the machine equipped with this monitor. It's robust, easy to use and perfectly matching for any environment. In a boat house or gym I don't need fuzzing around with any device to get some metrics, at home I can connect any BT device I want.
If I would be asked, my answer would be: change nothing.

[hikeplusrow]

I absolutely like the style how C2 provides the machine equipped with this monitor. It's robust, easy to use and perfectly matching for any environment. In a boat house or gym I don't need fuzzing around with any device to get some metrics, at home I can connect any BT device I want.
If I would be asked, my answer would be: change nothing.

Completely agree. The PM5 gives me all the metrics I need for an effective workout, and, like the rest of the machine, is completely reliable - increased complexity around tech would surely compromise this. Additionally, the RowErg, SkiErg and BikeErg are incredible value for money relative to other comparable cardio equipment - let's keep it that way.

[JaapvanE]

I suggested that C2 make a "Smart Tach" probably 5 years ago now.

Use a self powered Bluetooth tach with battery backup and just run the whole "Monitor" on your smartphone as an App, but Concept 2 have not been interested.

As explained in other threads: you physically can't. BLE is too slow and unreliable to do so. A PM5 has a lot of dataflow and calculations going on to get to the data it has. A general purpose OS like Android has too much going on to be able to reliably process the dataflow from a C2. With ORM we are close to hitting the 50% CPU load on a dedicated RPi, with all its cores set to max performance, and a Linux kernel set to PREEMPT. A phone simply can't process the data as it needs to do all other phone like stuff that is considered higher priority.

Bikesets essentially use a loadcell and some small buffering, which works well. In rowing some have experimented with this, and SmartRow is experimenting with a smart handle for comnercial use. Than, you don't need to do many complex calculations, but I have some reservations how they measure certain things well (linear displacement and timing are extremely complex as they have to integrate accuratly to get to some data).

With ORM we have a sister project that has partially taken that approach. It uses a cheap ESP32 chipset, do the data processing there and let a smartphone handle all presentation stuff. Although brilliantly implemented, you quickly run into issues that are unusable in a large market:

• Battery life requires a relatively weak CPU, thus CPU power is limited. As we found out, it is tough to get decent data processing working while not maxing out your CPU.
• Initial connect is quite a tough experience. People must connect in the blind as they don't have GUI feedback what they are doing. I don't see that working well when you have a fleet of machines where an eight crew starts at the same time.
• Everything is offloaded to the phone, and some stuff you can't, like ANT+. So you lose ANT+ capabilities, which are essential in gym scenarios where people want more detailed data.

RP3 and BioRower do preprocessing and use a tablet, but I think essentially they use a wired connection to get around a lot of the issues.

I think its coming, the PM6 will be your smartphone or tablet, people want the connectivity and a colour screen, especially if C2 take the real time loop and virtual rowing to the next level.

real time loop is a nice free toy, but EXR is now standard on the latest RP3 (with a 21" screen). Ergarta is also available now on C2. They missed the boat to add stuff. They are too far behind in the game and miss the user base. Might not be a bad thing: making a good high quality erg isn't easy, and as long as anybody connects to it, you're probably OK. Life might change when the SmartRow handle allows any machine to produce decent metrics.

C2 somehow need to offset the loss of money selling hardware and pull it in with software and a subscription service with some decent graphics. You get the basic monitor for free but you start paying for the add on's like the real time loop, a more pro version of the monitor that gives you performance analysis etc..

I don't think C2 loses money in selling C2's.

There are free games with quite a user base that look better than the real time loop. So there isn't much they can do here.

In all honesty: as soon as a company starts asking extra money for stupid stuff others provide for free (look at the RP3 data analysis they provide for free), they are in trouble. I won't buy a machine which needs a subscription to keep going (like Peleton or Hydrow) as I would volunteer it to become a hostage for a companies whims.

An add-on game, like EXR or Ergatta, I would pay for as they are an outside company. But C2 developing an EXR or Ergatta competitor is an extremely risky business which doesn't earn C2 extra credit with gym/boathouse owners. CoZweat seems to be dead in the water, despite being more real-life, the limitations of only using C2 amd iPads is too narrow. And lets forget the huge problems RowPeo currently has in getting a new version out the door. That also is the software business.

If I was C2, I'd stick to what I know, improve the PM5 a small bit, but not get into the software market.

[Carl Watts]

No way your phone is too slow over Bluetooth to process the raw data from a Bluetooth tach.

Way less content than music, not even close its just a changing frequency pulse stream , its about as basic a signal as you can get.

The PM5 is no longer simple, cheap or easy to service, its now a throw away device.

I just bought a new WattBike Atom X a couple of months ago, it comes with an Android tablet running the WattBike Hubb. Its the way to go, you want a big colour display. The PM5 is like something out of the 1980's, everyone has multiple devices these days from smartphones to tablets.

Maybe your local gym would want a PM5, but given a choice nobody at home would want it when you have a tablet or smartphone.

The PM5 is obsolete.

[JaapvanE]

No way your phone is too slow over Bluetooth to process the raw data from a Bluetooth tach.

A tach is a simple thing, a bicycle wheel doesn't rotate as fast, and it doesn't change speed as quickly.

A rowing flywheel rotates at least at 1000 RPM, and its speed variations are crucial to get metrics right. With 6 magnets you get a position report every 5 (drive) to 10 (recovery) ms. With high perfomance athletes this will drop to a position report every 3ms. So, at BLE's minimum 7.5 ms reporting interval, you'll get a position report way too late.

Getting the timing right is crucial in getting things like stroke detection and drag calculation to work. A phone will not allow an app to run a higher NICE levels and the standard kernels are not PREEMPT, and some other apps (like the entire phone drivers) have a much higher priority. So, a phone simply is incapable of processing hard real-time data. At ORM, 90% of our development effort is focused around making sure that we process our data in time, in an environment we complete control, including kernel parameters. We are hard-wired to the sensor as any latency in data processing in rowing has quite a determental effect on user experience.

Way less content than music, not even close its just a changing frequency pulse stream , its about as basic a signal as you can get.

You confuse latency with bandwith.

An audio stream from a phone to a headset can have some serious latency, but they fix it by adding pre-buffering and providing a lots of bandwith. You are only affected by latency when you change song unexpectedly.

A stream from an Erg is dominated by latency. You can't pre-buffer as you'll have no clue what is going to happen next. So while bandwith use is minimal, processing position reports too late (latency!) will cause havoc in stroke detection, drag calculation and force curves.

As it is fundamentally unknow what value comes next, you end up in post-buffering, combining several position reports in one message to get around the 7.5ms limit. Post-buffering has a dramatic effect on key feedback mechanisms, like the force curve, as you can't process it fast enough to display it realtime.

The PM5 is no longer simple, cheap or easy to service, its now a throw away device.

I think you have a totally misguided perception of the cost of SOC's. Retail price of a Raspberry Pi Zero 2W is €15, which runs ORM easily, including GUI, ANT+ and BLE. Major cost drivers are the battery, casing and screen. And it might be a "throw away device" for you, but in a gym/boathouse setup, you'll piss off the owner as a broken monitor will put an erg out of production, waiting for spare parts.

I just bought a new WattBike Atom X a couple of months ago, it comes with an Android tablet running the WattBike Hubb. Its the way to go, you want a big colour display. The PM5 is like something out of the 1980's, everyone has multiple devices these days from smartphones to tablets.

RP3 Aqua does the same. In all honesty, the monitor and app are the thing that age quickly. It is the screen quality and OS that move fast, making the entire device feel old quickly.

I think it is a rather wrong to fixate on the monitor itself. There are a lot of third-party apps that allow to placed over the monitor, making it less dominant in the user experience. The PM5 protocol is well equiped to allow a tablet to completely take over control of a PM5, including setting session parameters.

So, it then is up to the user which tablet he uses. I typically have EXR running, providing all essential metrics. I currently have a 15" Android tablet, but I'm considering upgrading to a 47" TV screen with a Nvidia Shield. A friend of mine has a projector, projecting to 120", giving a great emerging experience. Last thing you want to do is force a choice upon a user as it has to fit his home.

To me, it wiser to have a stable base monitor (PM5) and have a good option to add an external monitor that is easily replaceable every 4 years (C2 providing a tablet holder) and a decent protocol between them.

Maybe your local gym would want a PM5, but given a choice nobody at home would want it when you have a tablet or smartphone.

You realize that gym owners buy them by the dozens, right? The PM5 protocol is quite smart, allowing for quite some flexibility in usage scenarios. It is brilliant in its simplicity.

[JaapvanE]

Surely the device can take care of the extra BLE connections by rebroadcasting; the fancy graphics and metrics; the touch screen.

Rebroadcasting quickly becomes a total mess. It is a so much better user experience if all apps have a single source of truth. Technically, it isn't that complex to have two or three BLE connections. It is a software limitation in the PM5, not of the protocols used.

[stevegaspars]

It is a software limitation in the PM5

I'm guessing the PM5 doesn't have a lot of wriggle room for extra code to be run. CPU time would be a premium, it's not going to be of the same caliber as a Rpi. Multiple BLE connections will probably come at the sacrifice of something else. I once helped them track down a 'bug' in the firmware, where the code to merely operate the LED on the back of the monitor was affecting it's ability to measure stroke rate accurately under certain circumstances.

[SWE192]

Oh dear! I wasn't prepared for such opposing views. I certainly can see merit in all arguments.
My apologies for poking this wasp nest.

However, I am always a bit concerned about the general attitude of defending the status quo.

As an expert in corporate strategy and management, I have yet to see a company whose downfall hasn't started with a "what we have is good enough" attitude. I would hate to see Concept 2's long history of continuous improvement, among other things, reflected in the PM development break.

[Joebasscat]

I absolutely like the style how C2 provides the machine equipped with this monitor. It's robust, easy to use and perfectly matching for any environment. In a boat house or gym I don't need fuzzing around with any device to get some metrics, at home I can connect any BT device I want.
If I would be asked, my answer would be: change nothing.

This right here!