TrainingPeaks, the online training software that began as an interactive version of Joe Friel’s “TrainingBible” book series, is yet again singing in the updates shower. This time, in a wave of changes beginning in April and likely continuing through the end of June, Peaksware – makers of TrainingPeaks – sings the tune of HTML5 compatibility, so-called “artificial intelligence” features, and the ability to edit erroneous data. With some features still several weeks from rollout, Peakware invited VeloNews to their offices in Lafayette, Colorado, for a tour of the most recent updates.
iPhone-friendly Activity Viewer
The revamped Activity Viewer is the most significant of the current crop of changes. Released in May, it allows much more user-friendly workout viewing on any smartphone, which is a huge improvement.
Each workout has its own web address. Anybody with access either to the workout or its URL can view all critical workout information – graphs, metrics, map, and power profile – in one webpage. As the viewer is HTML5-compatible, anybody with access to the link or to the athlete’s calendar can open the workout on an iPhone or iPad.
This, and the ease with which an athlete or coach can access the viewer through the mobile app, finally makes TrainingPeaks useful on a mobile device.
Peaksware, in a blog post detailing the updates, depicts presumably a professional coach checking out clients’ workouts on his iPhone during a coffee break on one of his own training rides. Now that TrainingPeaks is genuinely functional on the small screen of an iPhone, that’s probably a realistic depiction.
One minor complaint is that it’s very easy to inadvertently cause the map, graph, or the peak power chart to scroll instead of the page itself. And if you’ve zoomed in on the graph, it may interpret your desire to scroll as a request to highlight data. After a few tries, though, it’s easy to find and avoid these error-prone points on the screen.
The Activity Viewer loads more slowly than we’re accustomed to on a pre-4G phone, but cyclists are a calm breed, and as far as the effect of a single change on the user experience, this is a big, big winner.
Predictive Performance Management Chart
The Performance Management Chart (PMC) could be a toy for most statisticians, but for athletes and coaches it’s a graph that illustrates current workload, fitness, and fatigue.
Based on the ability to estimate the training stress score (TSS) of future workouts, the chart now anticipates and displays future training load, fitness, and fatigue based on planned workouts. This helps to visualize their effect on fitness and fatigue, and should take some of the guesswork out of building fitness for self-coached athletes.
Because of the need to provide a planned TSS value for each individual workout, the predictive PMC would be most useful for planning a single block of training — for example, to maximize the quality of a taper in advance of an A-priority event, rather than a whole season.
It’s too soon to vouch for the accuracy of the chart’s foresight relative to real-world performances, but Peaksware says that its own athlete testing has shown the PMC’s predictive powers to be quite strong, and past personal experience suggests the same.
The data-editing feature in the map and graph workout page will be most handy for those with an SRM or PowerTap, or simply for those who forget to press the pause button. Users can now select and delete any segment of data that may have been gathered while walking or driving. There is now also the option to manually edit individual data points to remove random spikes or, if the mood strikes, to suddenly develop an 1800-watt sprint.
While certainly not critical to good data collection, the new functionality will allow for cleaner, clearer data, which should make analysis slightly easier, especially for coaches. Data editing is still in beta testing, but Peaksware expects to release it sometime in July.
Another addition is the automatic detection of power and heart-rate thresholds. TrainingPeaks will soon sweep a user’s uploaded data for 20- to 60-minute efforts, and immediately update either threshold if it detects a personal best effort. The software will then ensure that training load calculations for the PMC are brought up-to-date as well. It’s a handy feature that sends email notifications to the athlete and to a coach, requiring no action by the user other than the uploading of training files.
It’s too early to determine through beta testing if this will be accurate over the course of a season, but since it sweeps training files to determine new thresholds, it should be quite spot-on. Peaksware says that this is the first of what should be many upcoming intelligence features, most likely as part of their automated training-plan builder, and should be released in the next four to six weeks.
Peaksware’s invitation was primarily to demo the data-editing capability, but the HTML5 functionality for smartphones made the biggest impression. Both the threshold auto-detection and predictive PMC are useful additions that will take some of the trial-and-error out of self-coaching, while the data-editing capability will be mostly of interest to coaches who need hassle-free analysis.
Peaksware has said that its goal this summer is to improve the user experience, and the updates that have come since April go a long way towards meeting that objective.