This appears to be an EMI issue lacking system design details with SNR expectations, impedances, planned system design details, and cable types with proximity.
Line noise is common in sensor signals due to unbalanced impedance and crosstalk from noise in signals. Perhaps better-shielded cables on both pump power and sensor with better grounds, amplifier. Pumps can draw 5 to 10x the rated current starting at peak voltage, random phase AC starts can also induce spikes on signals both on turnon and turnoff, so a Zero Crossing Switched (ZCS) Triac in an SSR package can minimize the turn on and turn off transients and reduce the EMI.
AC?
If your "Motor Control" is simply random AC phase On/Off then yes a Triac SSR will be a big improvement. Improved cables and CM chokes can also improve EMI if you define the noise spectrum now and your requirements for air pressure Signal to Noise ratio. (SNR).
DC?
A DC motor control can reduce current peaks by RF PWM to start slower and this noise is easily filtered out once the noise levels are known and properly defined.
Bottom Line
- You must define the expected signal (level, BW, impedance)
- You must define the potential EMI noise sources and pump peak power requirements for any suitable switch. SSR's might not be able to handle an xxx hp pump for package offerings.
- You must estimate the required attenuation of noise ingress on signal to achieve your required SNR by shielding, and/or filtering isolation, ZCS, cable quality, orientation, etc.
- You can do this by pre-design testing components for the above requirements.
I had to do this many decades ago to learn what EMI looks like from all types of sources and how to reduce ingress effects on analog SNR. Dry contacts generate the broadest spectrum of noise. Even train locomotive electric engines radiate enough noise a block away onto CATV coax in the low and mid-bands from poor grounding to affect carrier to noise ratios in those frequency bands.
