Table 6.

Comparison of the effects of different actuation timings on the user’s effort during level walking

Ref.	Protocol	Onset timings	Peak timings	Offset timings	Metabolic cost	Soleus activity
					w.r.t. U	w.r.t. U
Galle, 2017 [21] ♢	4 onset conditions, fixed offset	Earliest (36%)	/	64%	-(14% - 18%)	-40% (peak) in Earliest, smaller reduction for later onset timings
		Early (42%)			-(16% - 21%)
		Late (48%)			-(16% - 17%)
		Latest (54%)			-8%
Malcolm, 2013 [18] ♢	5 onset conditions, fixed offset	13%	/	63%	-5%	/
		23%			-12%
		34%			-15%
		43%			-17%
		54%			-2%
Zhang, 2017 [46] □	Iterative learning to find optimal onset, peak, offset timings to reduce metabolic cost	17% - 37% (varied among subjects)	48% - 55% (varied among subjects)	59% - 65% (varied among subjects)	-(14% - 37%) (varied among subjects)	/
	or soleus activation	9%	44%	61%	/	-41% (rms)

The effects on the metabolic cost and soleus activity of healthy users during powered walking are reported with respect to the unpowered condition (U). The onset, peak and offset timings are expressed as a percentage of the gait time. In each study, the onset timings in bold are the values found to be the optimal ones to minimize the metabolic cost or the soleus activation in the subjects during walking

Symbols (♢,□) indicate studies performed by the same research group on similar actuation setups. Symbols are consistent between tables. In the works in which different experiments were performed (for example, different onset timings) the common information between the experiments (for example, same offset timing) is reported only once