2016 M2x MDH
2016 W2x POL
2016 M8+ St.Paul
2016 LM4- DEN
2016 W8+ Cornell

Evaluation of EmPower oarlocks

Evaluation of EmPower oarlocks

A quad was equipped with four EmPower oarlocks (star-board only) and the BioRowTel system (2). Two male junior crews consecutively performed the standard 2k BioRow test protocol with eight steps of the stroke rate (RBN 2013/03). The data was collected with both systems simultaneously, and then processed using the following methods: 

Samples of BioRowTel data were selected based on constant stroke rate (Fig.1) and processed using the standard averaging method in BioRow software. Then, derived values were calculated from averaged data for each sample and each rower (64 samples from two quads). Only data from the star-board was used and multiplied by two to match the data from EmPower oarlocks


Data from EmPower oarlocks was downloaded using LiNK for Windows software, exported into CSV format, then processed in MS Excel spreadsheets, where it was grouped into the same samples as BioRow data based on the constant stroke rate and corresponding number of strokes in each sample.

Tabl.1 presents correlations (n=64) of the data obtained using two systems, and the factor a in the linear regression y=ax between BioRow (x) and EmPower measurements (y):

The correlation was very high for all variables, except “Wash” (“Angle degrees traversed by the oarlock after the force drops below 100N”), where the regression was y=0.767x, which means the oarlock measure about 23% less “Wash” than the handle force sensor. This confirms our previous findings (RBN 2014/02) showing that the ratio of gate/handle forces is getting higher at the finish, so the “Wash” measured at the gate is shorter. This phenomenon could be related to inertia forces, but exact mechanics is still not completely clear to us. The lower correlation, also, could be explained by the difference in measurements methods: EmPower oarlock measure force at the gate in the horizontal direction only, but BioRow handle force sensor is mounted on the oar shaft, which changes orientation with oar squaring-feathering. At the finish, rowers may start feathering oar earlier or later, which affects BioRow measurements, but not EmPower data. 

Fig.2 shows that Total oar angle measured with EmPower oarlocks was about 6% longer than with BioRow sensor (the regression was y=1.06x), which was affected by the finish angle (12% longer in EmPower), while the catch angles were much closer (y=1.03x). This corresponds to our previous findings (RBN 2003/05) about the effect of a backlash of the oar in the oarlock for measurements of the finish angle.


Though correlation factor for the average force was slightly lower than for the angles, the values measured with EmPower oarlock were nearly directly proportional to the values measured with BioRow system (only -0.4% deviation of the regression slope from 1, Fig.3).


Rowing power measured with EmPower was also practically proportional to the values, measured with BioRow system (only 0.5% deviation, Fig.4).


Concluding, the most of biomechanical variables measured with EmPower oarlock directly correspond to BioRowTel measurements, and correction factors above should be applied for the Finish angle and “Wash”. Therefore, a large database of BioRow measurements could be reliably used for analysis and evaluation of EmPower data, which will be done in the next Newsletter.


1. NK EmPower Oarlock. http://www.nkhome.com/rowing-sports/empower-oarlock 

2. BioRowTel system. www.biorow.com 

©2017 Dr. Valery Kleshnev www.biorow.com